@article {pmid42481783,
year = {2026},
author = {Emekli, E and Demirel, BC and Demir, S and Yağcı, ZE and Oğuz, M and Tepe, M},
title = {Cloud-Enabled Automated CT Assessment of Pelvic Muscle Quality in Women With and Without Low-Energy Femoral Neck Fracture.},
journal = {Calcified tissue international},
volume = {117},
number = {1},
pages = {},
pmid = {42481783},
issn = {1432-0827},
mesh = {Humans ; Female ; *Femoral Neck Fractures/diagnostic imaging ; *Tomography, X-Ray Computed/methods ; Retrospective Studies ; *Muscle, Skeletal/diagnostic imaging ; Aged ; *Cloud Computing ; Middle Aged ; Aged, 80 and over ; *Pelvis/diagnostic imaging ; },
abstract = {This retrospective study aimed to compare pelvic muscle parameters between female patients with and without femoral neck fractures and to evaluate the feasibility of an automated CT-based workflow for exploring associations between pelvic muscle quality measures and femoral neck fracture status. The study included 119 female patients with low-energy femoral neck fractures and 107 age-matched female controls. Non-contrast computed tomography images were analyzed using a cloud-integrated deep learning segmentation tool. Pelvic muscle volume and quality parameters, including intramuscular adipose tissue, myosteatosis, and functional lean muscle, were quantified for the iliopsoas and gluteal muscles. Age-adjusted multivariate analysis, receiver operating characteristic analysis, and logistic regression were performed. Age and cortical bone parameters did not differ significantly between groups. In contrast, pelvic muscle quality parameters showed significant between-group differences after adjustment for age. Iliopsoas myosteatosis demonstrated the highest individual discriminatory performance, with an area under the curve of 0.711. The combined four-muscle myosteatosis model achieved the highest apparent within-sample discrimination, with an area under the curve of 0.739, indicating moderate discriminatory performance. Iliopsoas MYO values above the ROC-derived cut-off were associated with higher odds of belonging to the fracture group (OR, 4.22). Automated measurements showed excellent reliability. These findings suggest that pelvic muscle quality measures may provide complementary information for fracture-related assessment on routine CT examinations. Automated CT-based muscle quality assessment may provide a reproducible approach for opportunistic body-composition analysis on routine CT examinations. The cloud-enabled workflow used in this study illustrates technical feasibility; however, its clinical utility for fracture-risk stratification requires prospective validation.},
}

Humans
Female
*Femoral Neck Fractures/diagnostic imaging
*Tomography, X-Ray Computed/methods
Retrospective Studies
*Muscle, Skeletal/diagnostic imaging
Aged
*Cloud Computing
Middle Aged
Aged, 80 and over
*Pelvis/diagnostic imaging

@article {pmid42484935,
year = {2026},
author = {Schollmaier, P and Hohlmann, B and Wickel, N and Nguyen, HU and Yilmaz, O and Kovler, A and Feodorovici, P and Wellens, R and Radermacher, K and Janß, A},
title = {Toward ubiquitous surgical assistance services via private 5G infrastructure.},
journal = {International journal of computer assisted radiology and surgery},
volume = {},
number = {},
pages = {},
pmid = {42484935},
issn = {1861-6429},
support = {19OI22018C//Bundesministerium für Digitales und Verkehr (BMDV) - Federal Ministry for Digital and Transport of Germany/ ; },
abstract = {PURPOSE: Advanced data processing methods, including artificial intelligence (AI), are increasingly integrated into clinical practice, improving treatment quality and reducing practitioner strain. These methods entail elevated computational demands, necessitating secure and efficient resource provision. This study investigates a wireless infrastructure based on private 5G networking capable of provisioning scalable computational resources, exemplified by surgical assistance services (SAS) requiring live feedback.

METHODS: To this end, five exemplary SAS covering medical imaging, speech, and device data processing were deployed as virtualized applications on a private 5G-connected edge computing cluster. Latency and concurrent stream capacity were evaluated against a local processing baseline.

RESULTS: The private 5G configuration supports up to five concurrent 4K or 41 concurrent 720p video streams with on-site-like responsiveness and guaranteed latencies. A 720p end-to-end endoscopic SAS, covering local image capture, remote segmentation, and return, consistently achieves capture-to-display latencies below 290 ms.

CONCLUSION: The findings demonstrate how medical systems can wirelessly offload diverse computational workloads, satisfying responsiveness requirements for surgical applications despite a preliminary private 5G configuration. Compared to Wi-Fi, private 5G requires greater infrastructure effort but provides dedicated spectrum with reduced interference susceptibility, enhanced security, and temporal determinism, properties that may be of particular relevance for surgical navigation and robotic systems.},
}

@article {pmid42473943,
year = {2026},
author = {Li, Y and Yang, J and Hou, Q and Wang, S and Liu, S and Zhang, H and Ye, Z and Jiang, J and Lin, N and Xu, X and Shang, D and Qi, X and Wang, Z and Wang, H},
title = {Adaptive Redox Resistive Memory Programming for Efficient and Robust Class-Incremental Learning.},
journal = {Advanced materials (Deerfield Beach, Fla.)},
volume = {},
number = {},
pages = {e00025},
doi = {10.1002/adma.202600025},
pmid = {42473943},
issn = {1521-4095},
support = {2021ZD0201200//2030 Major Project of the Chinese Ministry of Science and Technology/ ; 62374181//National Natural Science Foundation of China/ ; U2341218//National Natural Science Foundation of China/ ; 92464201//National Natural Science Foundation of China/ ; 62488101//National Natural Science Foundation of China/ ; 62322412//National Natural Science Foundation of China/ ; XDA0330100//Strategic Priority Research Program of the Chinese Academy of Sciences/ ; 17212923//Hong Kong Research Grants Council/ ; C1009-22G//Hong Kong Research Grants Council/ ; C7003-24Y//Hong Kong Research Grants Council/ ; AOE/E-101/23-N//Hong Kong Research Grants Council/ ; T45-701/22R//Hong Kong Research Grants Council/ ; JLFS/E-601/24//Joint Laboratory of Microelectronics/ ; },
abstract = {Resistive memory (RM)-based computing-in-memory (CiM) accelerators provide a promising platform for low-power edge intelligence. However, practical edge AI requires not only efficient inference but also repeated model updates through class-incremental learning (CiL). Implementing CiL on RM substrates exposes a critical material-algorithm mismatch: the write-intensive updates required by CiL are strongly affected by the stochastic programming of filamentary RM devices. Conventional fully programming (FP) mitigates this stochasticity by repeatedly programming and verifying each cell to a precise target conductance, but this exhaustive procedure incurs substantial energy consumption, latency, and device wear across successive CiL stages. Herein, we propose a hardware-aware adaptive programming (AP) strategy that aligns CiL deployment with RM device physics. Microstructural and electrical analyzes reveal that the random spatial distribution of oxygen vacancies gives rise to unavoidable programming variability. Guided by this insight, AP does not attempt to eliminate intrinsic stochasticity through costly compensation. Instead, it updates only the most impactful weights to suppress accuracy loss caused by overall mapping errors, while leaving low-impact weights unchanged. This converts large-scale write-verify operations into targeted updates of a minimal subset of cells, reducing programming overhead without requiring device or material optimization. Validated on a hybrid analog-digital system with a 40 nm, 256 k RM-based CiM core, AP reduces programming energy by 93.0% and programming cycles by more than 90% relative to FP during five-stage CIFAR100 CiL, while achieving a final accuracy of 0.80, close to the 0.81 software baseline. For the more complex ShapeNet 3D point-cloud recognition task across eight stages, AP achieves 92.3% energy savings with only a 0.03 accuracy loss. Moreover, AP improves robustness by reducing programming-error-induced accuracy degradation by 87.7% and 88.4% for the two tasks, respectively. This work bridges algorithmic update requirements and physical programming constraints, enabling robust and energy-efficient lifelong learning on RM-based CiM platforms.},
}

@article {pmid42469293,
year = {2026},
author = {Harichandan, AK and Kumar, RR and Muduli, D and Mohapatra, SS and Parija, S and Kollie, DS and Priyadarshi, R and Ranjan, R},
title = {A requirement-driven framework for cloud service provider selection using AHP, QFD, and TOPSIS.},
journal = {Scientific reports},
volume = {},
number = {},
pages = {},
doi = {10.1038/s41598-026-61521-7},
pmid = {42469293},
issn = {2045-2322},
abstract = {Cloud service provider (CSP) selection has become a complex decision-making task as users must evaluate multiple functional requirements alongside competing quality-of-service (QoS) attributes. Although existing multi-criteria decision-making approaches can rank cloud services effectively, most lack a transparent mechanism for translating user requirements into measurable evaluation criteria. To address this limitation, this paper proposes a requirement-driven cloud service selection framework that integrates the Analytic Hierarchy Process (AHP), Quality Function Deployment (QFD), and Technique for Order Preference by Similarity to Ideal Solution (TOPSIS). The proposed framework first prioritizes user-defined functional requirements using AHP, subsequently maps these requirements to QoS attributes through QFD, and finally ranks CSP alternatives using TOPSIS based on requirement-aligned attribute weights. Experiments were conducted using real-world cloud service performance data comprising ten cloud service providers evaluated across five QoS attributes, namely response time, cost, security, scalability, and availability. The AHP phase produced highly consistent requirement priorities with a consistency ratio of 0.0074, while QFD-derived weights identified response time (0.264) and cost (0.204) as the most influential QoS attributes. The proposed framework consistently ranked CSP5, CSP9, and CSP2 as the most suitable providers. Comparative analysis against AHP-TOPSIS, MOORA, and VIKOR demonstrated strong ranking agreement with Kendall's Tau values exceeding 0.86, while Wilcoxon signed-rank tests confirmed statistically significant differences in ranking outcomes. Furthermore, sensitivity analysis verified ranking stability under alternative QFD relationship scales, and scalability experiments showed execution times below one second for scenarios involving up to 50 cloud service providers and 10 QoS attributes. The results demonstrate that the proposed framework offers an interpretable, robust, and computationally efficient solution for user-centric cloud service selection by explicitly linking user requirements to provider evaluation and ranking.},
}

@article {pmid42459140,
year = {2026},
author = {Taifa, IWR and Mperella, O},
title = {Enhancing Environmental Performance for Green Building Projects: Assessment Study due to the Applicability of Industry 4.0 Technologies.},
journal = {TheScientificWorldJournal},
volume = {2026},
number = {1},
pages = {e4976856},
pmid = {42459140},
issn = {1537-744X},
abstract = {The construction industry is embracing Industry 4.0 (I4.0) to improve environmental performance. This study thus evaluated the application of I4.0 on certified green buildings (GBs) projects. The study determined awareness level of I4.0 technologies on GBs, identified factors influencing I4.0 application, and proposed strategies for enhancing GBs' environmental performance. The results of the awareness assessment showed that, among all I4.0 technologies, artificial intelligence (AI) was the most well-understood. GBs apply some I4.0 technologies, including AI, Internet of Things, and cloud computing, and the awareness level for most I4.0-related technologies was found to be high, ranging from 2.68 to 4.68 (for the scale of 1-5). The awareness level of the GBs' principles was assessed, and the result showed that respondents were completely aware of two principles: energy efficiency and waste reduction. The overall mean score for the principles of GBs was 4.06, indicating a high level of awareness. The study also assessed the challenges of implementing I4.0, and noncritical challenges were a lack of standardization and certification, culture and aesthetics, client demands, and regulation and policy. The R-value was 0.723, indicating a high correlation. The R[2] = 0.523 shows how much the predictor variables technology, top management commitment, staff skills, infrastructure, financial arrangement, and operational control factor can account for the variation in the environmental performance. Consequently, the study found perception-based projections rather than outcomes verified through objective measurements that the I4.0 technologies are projected to enhance energy efficiency (15%-25%), water conservation (≥ 20%), improve indoor environmental quality, improve material selection and lead to waste reduction. Therefore, I4.0 technologies can enhance GBs' performance by enabling real-time monitoring and data-driven optimization. I4.0 facilitates energy and resource management through smart systems, extends system lifecycles via predictive maintenance, and supports better environmental impact assessments in construction design. Despite challenges like high costs and cybersecurity risks, I4.0 technologies apply to both new and existing buildings, enabling dynamic, evidence-based decision-making that significantly improves environmental metrics and operational efficiency.},
}

@article {pmid42460000,
year = {2026},
author = {Nwadinigwe, KJ and Oduoye, MO and Jehan, M and Shahid, A and Hassan, A and Mudassir, SS and Bashir, U},
title = {Exploring an AI-driven dynamic triage system for real-time patient risk reassessment in emergency departments in low-resource settings.},
journal = {Frontiers in digital health},
volume = {8},
number = {},
pages = {1776341},
pmid = {42460000},
issn = {2673-253X},
abstract = {BACKGROUND: Emergency department (ED) overcrowding is a global challenge, particularly acute in low-resource settings, where staff and equipment shortages exacerbate inefficiencies. Conventional triage systems are static and often fail to detect patient deterioration after the initial assessment. This study aims to explore the potential of AI-driven dynamic triage systems for continuous reassessment of patient risk in low-resource EDs.

METHODS: A narrative review was conducted through a literature search from 2014 to 2025, following a structured selection process using PubMed, Google Scholar, Web of Science, and ResearchGate, synthesizing evidence on conventional triage models, AI- based decision support, and machine learning applications in emergency care.

RESULTS: This study shows that the Emergency Severity Index (ESI) achieves a pooled sensitivity of 81.8% and specificity of 70.5-81.7 for predicting short-term mortality and ICU admissions; yet, under-triage rates in LMICs reach 25%-30%. AI-driven models, including logistic regression, random forests, gradient boosting, and LSTM networks, demonstrate superior predictive accuracy, reducing prioritization errors by up to 9% and improving detection of high-risk patients. Pilot programs in LMICs confirm the feasibility of mobile and cloud-based AI tools, though challenges remain in data quality, infrastructure, and clinician trust.

CONCLUSION: AI-enabled dynamic triage offers promise for enhancing patient safety, optimizing resource allocation, and reducing mortality in overcrowded EDs. However, successful implementation requires prospective validation, infrastructure investment, and co-design with clinicians to ensure adaptability in low-resource settings.},
}

@article {pmid42463256,
year = {2026},
author = {Moriya, T},
title = {GoToCloud: Development and future of the cloud-based platform for cryo-EM structure-based drug DesignDevelopment and future of GoToCloud.},
journal = {Progress in molecular biology and translational science},
volume = {223},
number = {},
pages = {203-233},
doi = {10.1016/bs.pmbts.2026.03.006},
pmid = {42463256},
issn = {1878-0814},
abstract = {The "Resolution Revolution" in cryogenic electron microscopy (Cryo-EM) has fundamentally advanced structure-based drug design (SBDD), moving beyond static snapshots to dynamic conformational analysis. However, the resulting exponential growth in data volume has created critical computational bottlenecks where traditional on-premise high-performance computing lacks necessary scalability and economic flexibility. This chapter describes "GoToCloud," a platform leveraging Amazon Web Services (AWS) ParallelCluster to overcome these barriers. We introduce a "Shared EFS" architecture and automated scripts allowing researchers to deploy secure virtual clusters within a Virtual Private Cloud (VPC), pre-installed with essential software like RELION. This design eliminates specialized IT maintenance needs for researchers using Cryo-EM while ensuring consistent analysis environments across distributed groups. Extensive benchmarking using practical datasets, including Nitrite Reductase and Streptavidin, validates the platform's effectiveness. Crucially, we provide a detailed economic analysis optimizing cost-performance by identifying specific GPU instance types (e.g., NVIDIA T4 vs. A10G) that balance processing speed with expense. The platform achieved 1.83Å resolution, revealing the "2.0Å wall" for cost-performance. Additionally, the platform addresses global data logistics by integrating Zettar zx, a high-speed transfer technology utilizing burst-buffer architecture. This successfully demonstrated ∼4.6 Gbps throughput over intercontinental distances despite 10 Gbps on-premise bottlenecks, replacing slow physical media transport. By democratizing access to High-Performance Computing (HPC) and enabling rapid data mobility, GoToCloud establishes a robust foundation for next-generation SBDD, facilitating fully automated high-throughput screening and experimental validation of AI-predicted models.},
}

@article {pmid42463815,
year = {2026},
author = {Jayanthi, V and Sivakumar, S},
title = {Edge optimized hybrid quantum-classical ensemble framework for EEG and MRI based epileptic seizure detection in IoMT.},
journal = {Scientific reports},
volume = {},
number = {},
pages = {},
doi = {10.1038/s41598-026-60687-4},
pmid = {42463815},
issn = {2045-2322},
abstract = {Accurate and timely detection of epileptic seizures is critical for clinical intervention however, conventional EEG analysis is constrained by noise, inter-patient variability, and high computational demands. To overcome this the proposed lightweight Hybrid Quantum classical Ensemble Net (HQCE-Net) optimized for edge device. The proposed pipeline with filtering, noice reduction scaling to suppress artifacts, followed by extraction of features on EEG sub-band powers (δ, θ, α, β) and MRI images multi model feature fusion using Quantum Fourier Transform (QFT) and Quantum Wavelet Transform (QWT) enabling complementary frequency-domain and multi-resolution time-frequency representation while maintaining computational efficiency. The fused features are standardized using z-score normalization. The complete system is implemented on a Raspberry Pi 5, On the EEG dataset (CHBMIT), HQCE-Net achieves about 92% accuracy with 92% performance metrics. On the MRI dataset it achieves 98% accuracy with 98% precision/recall, showing strong and consistent performance across both EEG and MRI modalities on edge device. Unlike existing EEG seizure detection methods that rely on computationally intensive deep models or cloud-based processing, This work demonstrates the practical and reliable neurodiagnostic monitoring for home-based and resource-limited clinical environments.},
}

@article {pmid42450765,
year = {2026},
author = {Deng, X and Zhang, F and Jin, G and Cui, L and Zhang, D and Zhang, F},
title = {Recent Advances in Vision-Based Beef Cattle Body Measurement Technologies.},
journal = {Animals : an open access journal from MDPI},
volume = {16},
number = {13},
pages = {},
doi = {10.3390/ani16132058},
pmid = {42450765},
issn = {2076-2615},
support = {2022KJ126//Scientific Research Project of Tianjin Municipal Education Commission/ ; 2025TJMT036//Tianjin Metrology Science and Technology Project/ ; 2025YFDZ0108//Key R&D and Achievement Transformation Program of Inner Mongolia Autonomous Region/ ; },
abstract = {Accurate beef cattle body measurement data are crucial for growth assessment, phenotypic analysis, breeding management, and precision livestock farming. Traditional manual measurements are labor-intensive, time-consuming, and likely to cause stress in animals, making it difficult to meet the demands of large-scale livestock farming. This paper employs a structured systematic literature review method, in accordance with the PRISMA 2020 guidelines, to summarize research progress in vision-based beef cattle body measurement. This paper focuses on reviewing technical approaches such as 2D image-based measurement, 3D measurement using RGB-D and LiDAR, and multi-view fusion. It analyzes key technologies including image segmentation, keypoint detection, point cloud processing, 3D reconstruction, and geometric calculations, and compares the advantages and disadvantages of different methods in terms of measurement accuracy, robustness, cost, and farm applicability. The results indicate that 2D image-based methods are low-cost and flexible to deploy but have limited expressiveness for 3D body measurement parameters; RGB-D and LiDAR methods can provide spatial information but are affected by point cloud noise, occlusion, equipment costs, and data processing complexity; multi-view fusion can improve the completeness of body surface information but places high demands on calibration, registration, and system integration. Current research still faces challenges such as a lack of public datasets, inconsistent annotation standards, uncertainty regarding ground truth, insufficient cross-ranch generalization validation, and limited practical applications. Future research should focus on developing standardized datasets, conducting cross-scenario validation, advancing multimodal perception, creating lightweight models, and applying edge computing to drive the evolution of visual body measurement toward continuous monitoring and intelligent decision-making.},
}

@article {pmid42451307,
year = {2026},
author = {Torad, MA and Torad, AAM and Taha, MM and El-Mokadem, ES},
title = {IoT-Based Isolation Ward Monitoring System Prototype.},
journal = {Sensors (Basel, Switzerland)},
volume = {26},
number = {13},
pages = {},
doi = {10.3390/s26134065},
pmid = {42451307},
issn = {1424-8220},
support = {PNURSP2026R209//Princess Nourah bint Abdulrahman University/ ; },
mesh = {Humans ; COVID-19 ; *Internet of Things ; SARS-CoV-2 ; Pandemics ; Remote Patient Monitoring ; Convolutional Neural Networks ; Electrocardiography ; Digital Health ; Monitoring, Physiologic/methods/instrumentation ; Oximetry ; Betacoronavirus ; *Pneumonia, Viral ; *Coronavirus Infections ; Dashboard Systems ; },
abstract = {The COVID-19 pandemic exposed critical vulnerabilities in healthcare systems worldwide, placing healthcare workers (HCWs) at severe infection risk through direct patient contact. Epidemiological data confirm that HCWs were approximately seven times more likely to develop severe COVID-19 than other occupations, with over 7000 HCW deaths recorded globally by mid-2020. This paper presents the design and laboratory proof-of-concept validation of an IoT-based remote patient-monitoring system prototype-the IoT-Based Isolation Ward Monitoring System Prototype-designed to eliminate unnecessary patient-to-HCW physical contact while maintaining continuous, real-time physiological surveillance. The system integrates multi-sensor hardware comprising an AD8232 ECG module, a MAX30100 pulse oximeter, an NTC thermistor, and an MQ-135 CO2 sensor. These sensors interface with an Arduino UNO for data acquisition, while localized edge computing is executed on a Raspberry Pi 3B. A convolutional neural network (CNN) trained on the MIT-BIH Arrhythmia Database classifies heartbeats into five distinct categories. By utilizing SMOTE resampling on 109,446 samples, the network achieves an on-device inference latency of under 200 ms. The sensor data are transmitted to a Firebase Realtime Database via an authenticated REST API, which synchronizes data across dual front-end interfaces: a LabVIEW desktop dashboard for clinical oversight and a cross-platform Flutter mobile application for mobile monitoring. End-to-end technical validation under controlled laboratory conditions confirmed round-trip cloud latencies between 300 and 800 ms, error-free threshold alert generation, and sub-second latency for the integrated chat utility. The proposed system uniquely combines hardware sensing, ML-based ECG classification, cloud storage, a LabVIEW physician dashboard, and bidirectional doctor-patient mobile communication into a single unified, low-cost platform.},
}

Humans
COVID-19
*Internet of Things
SARS-CoV-2
Pandemics
Remote Patient Monitoring
Convolutional Neural Networks
Electrocardiography
Digital Health
Monitoring, Physiologic/methods/instrumentation
Oximetry
Betacoronavirus
*Pneumonia, Viral
*Coronavirus Infections
Dashboard Systems

@article {pmid42451359,
year = {2026},
author = {Qadri, JA and Sajjad, A and Khan, AHA},
title = {Spatiotemporal Modeling of Mangrove Carbon Stock Along Pakistan's Coast Using Multi-Sensor Sentinel and Landsat Data.},
journal = {Sensors (Basel, Switzerland)},
volume = {26},
number = {13},
pages = {},
doi = {10.3390/s26134117},
pmid = {42451359},
issn = {1424-8220},
abstract = {This study quantifies coastal mangrove carbon stocks and their interannual variability along the Pakistan coastline by developing a multi-sensor fusion framework integrated with a process-based light use efficiency (LUE) modeling approach. To ensure high-cadence monitoring and overcome persistent cloud cover over the Indus Delta, data from multiple satellite sensors including Landsat 8/9 and Sentinel-2 within Google Earth Engine were utilized. Sentinel-2-derived Normalized Difference Vegetation Index (NDVI) data composited for the January-March period was processed to estimate vegetation productivity. Field-based validation of biomass estimates was conducted using 57 georeferenced sampling points, cross-compared with Sentinel-2 data. Mangrove extent was delineated through land use and land cover (LULC) classification into water bodies, mangroves, mudflats, land parcels, and sand surfaces. The LUE model incorporated environmental stress scalars, including temperature, vapor pressure deficit (VPD), salinity, and photosynthetically active radiation (PAR) to estimate gross primary productivity and derive total biomass, which was subsequently converted into carbon stocks. Results indicate a mean carbon stock of 31.95 Mg C ha[-1] (equivalent to 117.3 Mg CO2 ha[-1]), with significant interannual variation (coefficient of variation = 19.8%). A significant decline in carbon stocks was observed in 2021 (-11.11%; 3.56 Mg C ha[-1]), corresponding to a reduction in NDVI value (0.55 compared to 0.58 in other years). Spatial analysis revealed substantial heterogeneity in carbon distribution (20.51 to 55.93 Mg C ha[-1]), primarily influenced by localized salinity gradients and water stress conditions. This study mapped mangrove extent, quantified environmental stress, and estimated carbon stocks across Pakistan's coast from 2020 to 2024, delivering a spatially resolved, multi-year baseline for coastal carbon assessment and ecosystem monitoring in arid tidal environments.},
}

@article {pmid42451531,
year = {2026},
author = {Wang, Z and Chen, J and Zhao, H and Wei, B},
title = {Deterministic Edge-Controlled Precision Fertigation System with Spatial Task Scheduling and Hardware-Software Safety Interlock.},
journal = {Sensors (Basel, Switzerland)},
volume = {26},
number = {13},
pages = {},
doi = {10.3390/s26134289},
pmid = {42451531},
issn = {1424-8220},
abstract = {Cloud-dependent irrigation platforms can support remote monitoring, but their use in precision fertigation is limited when local decisions must be made quickly and reliably. Network delay, temporary disconnection, and the use of single-point measurements may all reduce the ability of a system to respond to spatial variation in soil moisture and nutrient demand. In this work, an edge-controlled precision fertigation system was developed by combining multi-parameter soil sensing, spatial task scheduling, and a 6-DOF robotic manipulator. The ESP32 controller runs a preemptive FreeRTOS scheduler, allowing sensor acquisition, inverse-kinematics calculation, and pump actuation to be handled as separate tasks. A Kalman filter was used to smooth soil moisture measurements, and a hysteresis-based control strategy was adopted to reduce false triggering and repeated pump switching. To improve fertigation safety, a hardware-software interlock was added so that fertilizer delivery is always accompanied by water delivery. Hardware-in-the-Loop simulation and a 14-day field deployment were used to evaluate the system. The controller achieved an end-to-end latency of less than 38 ms and maintained operation during network interruptions through cached local parameters. After calibration, the robotic end-effector positioning error was reduced to ±2.4 mm. The hysteresis strategy lowered daily pump cycling by 71%. Based on prototype duty-cycle data and seasonal extrapolation, the projected seasonal water use and fertilizer demand were 44% and 38% lower, respectively, than those estimated for a uniform application. These values should be interpreted as model-based projections rather than direct season-long measurements. During 72 h of continuous operation, no Modbus faults were observed, and RTOS heap fragmentation remained stable. Overall, the results suggest that edge-based deterministic control can provide a practical route for precision fertigation where both spatial variability and intermittent connectivity must be considered.},
}

@article {pmid42451558,
year = {2026},
author = {Čilić, I and Podnar Žarko, I and Kušek, M and Štajdohar, J},
title = {Cost-Aware Scheduling Under Latency Constraints for Multi-View 3D Reconstruction Across the Edge-Cloud Continuum.},
journal = {Sensors (Basel, Switzerland)},
volume = {26},
number = {13},
pages = {},
doi = {10.3390/s26134317},
pmid = {42451558},
issn = {1424-8220},
support = {NPOO.C3.2.R3- I1.04.0070//European Union/ ; },
abstract = {Learning-based multi-view 3D reconstruction pipelines, such as transformer-based approaches, enable the accurate reconstruction of 3D scenes from multiple images, but their deployment across the edge-cloud continuum is challenging due to high computational demands and large intermediate data transfers. Effective pipeline scheduling in the continuum must therefore balance latency constraints with the cost of cloud resource usage. In this work, we address cost-aware scheduling under latency constraints for a multi-stage 3D reconstruction pipeline consisting of depth estimation, transformer-based multi-view fusion, and point cloud merging with export to a rendering-ready representation. We implement a service-oriented pipeline where each stage can be executed either on edge or cloud nodes, and we experimentally characterize its performance on representative hardware platforms. The results show a strong imbalance between the computational time and communication latency across platforms, mainly due to large intermediate data. Based on these insights, we propose an online scheduler that dynamically selects stage placements to minimize the cloud cost while satisfying latency constraints. The scheduler incorporates a top-K edge selection mechanism that reduces the decision complexity by jointly considering the network conditions and node utilization. Simulation results parameterized with real-system measurements show that the proposed approach effectively reduces cloud usage while meeting latency constraints, outperforming the baseline strategies based on single-node pipeline execution.},
}

@article {pmid42457743,
year = {2026},
author = {Javadpour, A and Taleb, T and Benzaid, C and Zeraoulia, K and Djebrani, A and Belhadi, MY and Cordeiro, L and Rosa, L},
title = {Cloud-native encryption as a service for IoT.},
journal = {Scientific reports},
volume = {16},
number = {1},
pages = {},
pmid = {42457743},
issn = {2045-2322},
abstract = {Encryption as a Service (EaaS) is a practical solution for resource-constrained Internet of Things (IoT) devices that cannot efficiently execute costly cryptographic tasks locally. This paper presents a cloud-native EaaS platform implemented on Kubernetes and designed to support scalable encryption, decryption, and key-management services for IoT environments. The paper describes the functional architecture of the platform, defines its main service workflow, and introduces two deployment modes, namely cloud-based and fog-based deployment. The proposed platform is evaluated in terms of processing time, deployment time, and end-to-end response time. The results show that the fog-based deployment reduces the response time by at least [Formula: see text] for small payloads and by up to [Formula: see text] for larger payloads compared with the cloud-based mode. The deployment analysis also shows that increasing the number of replicas from 1 to 5 leads to a deployment-time increase of more than [Formula: see text], while increasing the workload to 11 replicas results in an increase of about [Formula: see text]. In addition, the results indicate that the Key Manager is the most resource-intensive component and has the highest impact on pod readiness time. Overall, the findings show that the proposed Kubernetes-based EaaS platform can provide flexible and scalable cryptographic support for IoT systems, while fog-based placement offers clear latency advantages in the evaluated prototype setting.},
}

@article {pmid42442246,
year = {2026},
author = {Łukowski, A and Papier, G and Wójcik, R and Domżał, J},
title = {A survey of the integration between machine learning and artificial intelligence techniques in software-defined networking.},
journal = {Neural networks : the official journal of the International Neural Network Society},
volume = {205},
number = {Pt A},
pages = {109339},
doi = {10.1016/j.neunet.2026.109339},
pmid = {42442246},
issn = {1879-2782},
abstract = {Recently, Software-Defined Networking has become one of the key technologies in modern telecommunication networks. Its programmable and flexible architecture enables centralized control of network resources, automation of management processes, and integration with other technologies such as Network Function Virtualization, Edge/Fog/Cloud Computing, and 5G/6G networks. Moreover, SDN provides a natural environment for implementing advanced Machine Learning and Artificial Intelligence techniques, which are increasingly applied to anomaly detection, network traffic analysis, load prediction, and flow optimization. This article presents a comprehensive survey of the integration between SDN and AI across diverse networking domains, including core, cloud, edge, wireless, IoT, and vehicular networks. More than 400 publications from 2015 to 2025 have been analyzed and classified according to network domain, application area, and category of AI techniques employed-particularly supervised, unsupervised, and hybrid learning methods. Based on this systematic analysis, a mapping framework was developed to illustrate the relationships between AI techniques and SDN application domains, revealing dominant research trends and existing gaps in the current literature. Unlike many previous surveys that focus on individual SDN aspects, selected network domains, or isolated AI-based applications, this survey provides an integrated cross-domain perspective on AI-enabled SDN. Its main contribution lies in a taxonomy-driven mapping of machine learning paradigms, SDN application areas, and network domains, which clarifies how current research supports the transition from programmable SDN toward more autonomous and intelligent network management. The results of the conducted review indicate that the integration of SDN and AI constitutes a foundation for the development of intelligent network management mechanisms. The machine learning techniques presented in this paper form the basis of the currently emerging AI-driven solutions for SDN environments, supporting prediction, classification, anomaly detection, and network performance optimization. At the same time, it should be emphasized that the supervised, unsupervised, and hybrid paradigms primarily address prediction, classification, detection, and adaptive optimization functions, whereas full network automation also requires interactive learning mechanisms, continuous decision-making, multi-domain orchestration, and closed-loop control. In the final part of the article, the main directions for future research related to the development of more autonomous, adaptive, and explainable network management architectures are also identified.},
}

@article {pmid42432137,
year = {2026},
author = {D, AS and Hegde, A and C M, AC and V, N and H M, A},
title = {Performance evaluation of TCP congestion control variants across application workloads in cloud based networks.},
journal = {Scientific reports},
volume = {},
number = {},
pages = {},
doi = {10.1038/s41598-026-61100-w},
pmid = {42432137},
issn = {2045-2322},
abstract = {Choosing the right Transmission Control Protocol (TCP) congestion control algorithm matters more in shared cloud environments than is often appreciated, yet head-to-head comparisons across heterogeneous, concurrently running workloads remain rare in the literature. We evaluated three widely used variants Cubic, Reno, and Bottleneck Bandwidth and Round-trip propagation time (BBR) under four workload types: synthetic throughput testing with iperf3, real-time stream processing using Apache Kafka and Apache Flink, distributed I/O, and compute-intensive sorting via Hadoop TeraSort. All experiments ran inside a dumbbell network topology built on four Amazon Web Services (AWS) m7i-flex.large Elastic Compute Cloud (EC2) instances, with a dedicated forwarding node and static routing forcing every flow through that single contention point. Across every metric we measured throughput, end-to-end latency, retransmission count, and job completion time the three variants behaved quite differently depending on the workload. The results give concrete, workload-specific guidance for operators choosing a congestion control policy in multitenant cloud deployments.},
}

@article {pmid42433908,
year = {2026},
author = {Zhang, J and Peng, Y and Wang, Z and Yu, W and Zhong, C and Yang, C and Liu, Y and Cao, G},
title = {Predicting the distribution of Deyeuxia angustifolia habitats in the Tumen River Basin due to climate and land-use changes.},
journal = {Frontiers in plant science},
volume = {17},
number = {},
pages = {1865223},
pmid = {42433908},
issn = {1664-462X},
abstract = {Habitat dynamics of the representative wetland plant, Deyeuxia angustifolia serve as a key indicator of ecosystem health in the transboundary Tumen River Basin. To investigate spatiotemporal evolution under combined climate and land-use changes, we integrated multi-source environmental variables into a species distribution model (SDM) using the Google Earth Engine (GEE) cloud platform. Current habitat suitability was simulated, and distributions for the 2050s were projected under three shared socioeconomic pathways (SSP126, SSP245, and SSP585). Currently, high and moderate suitability habitats cover 25.5% of the basin, concentrated in Russia's khasan District, China's Wangqing County, and Ryanggang Province in the Democratic People's Republic of Korea. Future scenarios suppress distribution; however, as emission concentrations rise, habitat loss areas decreased, with land-use change emerging as the primary driver of expansion. Habitat changes were most pronounced in China, under the low-emission SSP126 scenario. Conversely, under SSP585, suitable areas remain the largest among future projections, though still below current baselines. This study coupled climatic and anthropogenic variables and addressed previous modeling limitations. The study findings provide scientific support for mitigating wetland degradation, conserving biodiversity, and guiding ecological management in the Tumen River Basin.},
}

@article {pmid42436157,
year = {2026},
author = {Yao, J and Haixiang, G and Yanping, C and Qizhi, Z and Almenweer, RA and Isleem, HF and Mostafa, AM},
title = {Cloud-enabled hybrid structural equation modeling and artificial neural network framework for energy-efficient green buildings.},
journal = {Scientific reports},
volume = {},
number = {},
pages = {},
doi = {10.1038/s41598-026-60114-8},
pmid = {42436157},
issn = {2045-2322},
abstract = {Rapid urbanization and the growth in demand for sustainable infrastructure have made energy efficiency in the built environment a world priority. Nonetheless, the inherent gap between predictions and actual operational use is a fundamental challenge, as linear models, almost exclusively, often fail to model the complexity of non-linear relationships between climatic conditions, building physics, and dynamic occupancy. To overcome these shortcomings, this study presents the development of a novel approach in the form of a hybrid scheme that combines Structural Equation Modeling with an advanced approach using Artificial Neural Network and its reinforcement through a distributed cloud computing architecture. The proposed approach uses a multi-head attention mechanism to dynamically weight the importance of features and uses cloud-based load balancing to efficiently process high-dimensional datasets. Empirical results show that this model provides an excellent fit to the empirical data, yielding a coefficient of determination ([Formula: see text]) of 0.885 on the training set and 0.879 on the testing set, outperforming conventional forecasting benchmarks. While individual small-scale consumption outliers yield an arithmetic Mean Absolute Percentage Error (MAPE) of 112.4%, the volume-weighted MAPE (WMAPE) is kept at a highly stable 10.3%, confirming robust absolute precision. Additionally, distributed cloud simulations demonstrate an eight-fold increase in processing speed alongside an average computational energy saving of 12.16 kWh per processing cycle. These findings validate that merging artificial intelligence with scalable cloud architectures enables viable real-time energy optimization, providing stakeholders with a reliable blueprint to reduce carbon footprints and achieve long-term sustainability.},
}

@article {pmid42416581,
year = {2026},
author = {Chen, Y and Yu, Q and Xu, H and Zhang, L and Li, W and Wu, Y},
title = {Innovative Advances in Non-Invasive Detection Technologies for Heart Failure: Synergistic Application of Multimodal Sensing and Intelligent Algorithms.},
journal = {Reviews in cardiovascular medicine},
volume = {27},
number = {6},
pages = {48196},
pmid = {42416581},
issn = {2153-8174},
abstract = {Heart failure (HF) remains a leading global cause of chronic disease-related disability and mortality, with rising incidence driven largely by population aging. Early diagnosis is challenging because initial symptoms are often subtle and non-specific, leading to delayed detection and poor prognosis. While conventional tools such as echocardiography and B-type natriuretic peptide (BNP) testing remain diagnostic gold standards, these approaches are limited by operator dependency, restricted accessibility, and dynamic monitoring. Recent advances in artificial intelligence (AI) and cloud computing have enabled a new generation of non-invasive, intelligent technologies that integrate wearable sensors (e.g., ReDS™) with multimodal platforms (e.g., HeartLogic™, CardioSignal) to support real-time risk tracking and personalized management. Indeed, supported by favorable policy environments and strengthened collaboration among manufacturers, clinicians, and researchers across multiple fields and disciplines, the development of intelligent non-invasive HF detection devices has accelerated, leading to rapid innovation, commercialization, and continuous emergence of novel technologies and products. This review systematically summarizes HF pathophysiological mechanisms and current clinical monitoring strategies. Moreover, this review critically evaluates emerging devices and AI-driven platforms, highlighting the associated underlying principles, data integration capabilities, and clinical applicability. Finally, the analysis addresses key challenges, including the "black box" dilemma associated with AI, data bias, and privacy concerns, and proposes future directions for early screening, risk stratification, and precision intervention. By synthesizing technological comparisons and limitations, this review aims to provide a comprehensive reference for advancing intelligent HF diagnostics.},
}

@article {pmid42417454,
year = {2026},
author = {Liu, Y and Yan, K and Ma, S and Deng, H and Kong, J},
title = {AI-Driven Digital Twin Architecture for Multimodal Prediction and Adaptive Intervention in Cognitive Aging.},
journal = {JMIR AI},
volume = {5},
number = {},
pages = {e87768},
doi = {10.2196/87768},
pmid = {42417454},
issn = {2817-1705},
abstract = {Age-related cognitive dysfunction, including mild cognitive impairment and dementia, underscores the need for scalable and personalized predictive models. We present a conceptual artificial intelligence-driven digital twin framework to support early detection, real-time monitoring, and adaptive intervention. The system is structured around 4 core processes: perception, analytics, decision-making, and adaptive feedback, and is organized across 5 functional layers: data acquisition, integration, modeling, reasoning, and application. Multimodal behavioral, physiological, and clinical data are harmonized using Fast Healthcare Interoperability Resources and Observational Medical Outcomes Partnership standards. Predictive modeling uses convolutional and recurrent neural networks, gradient boosting, and reinforcement learning. The framework is designed for cloud-based deployment on platforms that support HIPAA-aligned implementation, including Amazon Web Services and Microsoft Azure, with 7 application modules spanning signal-based and pose-based assessment, personalized mind-body training, cognitive rehabilitation, and disease trajectory simulation. This architecture offers a foundation for precision cognitive care in aging populations.},
}

@article {pmid42418520,
year = {2026},
author = {Lee, D and Park, J and Lee, J and Lee, C and Park, YP},
title = {Compression benchmarking of holotomography data using OME-Zarr format.},
journal = {PloS one},
volume = {21},
number = {7},
pages = {e0351560},
doi = {10.1371/journal.pone.0351560},
pmid = {42418520},
issn = {1932-6203},
mesh = {*Data Compression/methods ; Benchmarking ; Algorithms ; Compression Algorithms ; *Imaging, Three-Dimensional/methods ; Image Processing, Computer-Assisted/methods ; Animals ; },
abstract = {Holotomography (HT) is a label-free, three-dimensional quantitative phase imaging technique that captures refractive index distributions of biological samples at sub-micron resolution. As modern HT systems enable high-throughput and large-scale acquisition, they produce terabyte-scale datasets that require efficient data management. This study presents a systematic benchmarking of data compression strategies for HT data stored in the OME-Zarr format, a cloud-compatible chunked data structure suitable for scalable imaging workflows. Using six representative datasets from five biological samples, we evaluated combinations of preprocessing filters and 13 compression algorithms across multiple compression levels. Performance was assessed in terms of compression ratio, bandwidth, and decompression speed. A throughput-based evaluation metric was introduced to capture realistic performance under varying network constraints, revealing that the optimal compression strategy is strongly dependent on available system bandwidth. Across a wide range of bandwidth conditions, Pcodec consistently exhibited the most balanced overall performance, followed by Blosc-zstd and zstd. The results offer practical guidance for the storage and transmission of large HT datasets and serve as a reference for implementing scalable, FAIR-aligned imaging workflows in cloud and high-performance computing environments.},
}

*Data Compression/methods
Benchmarking
Algorithms
Compression Algorithms
*Imaging, Three-Dimensional/methods
Image Processing, Computer-Assisted/methods
Animals

@article {pmid42414419,
year = {2026},
author = {Alnafisah, KH and Almutairi, AM and Ibraheem, A and Almutawa, A and Bajahzar, AS and Nasser, T and Alyahya, AN},
title = {Scalable hierarchical federated graph-transformer architecture for efficient multi-modal intrusion detection in 6G UAV-assisted vehicular IoT.},
journal = {Scientific reports},
volume = {},
number = {},
pages = {},
doi = {10.1038/s41598-026-60184-8},
pmid = {42414419},
issn = {2045-2322},
abstract = {The increasing growth of 6G-empowered UAV-assisted vehicular IoT systems brings forth unprecedented scalability issues for distributed intrusion detection, especially in the context of non-IID data distributions and heterogeneous edge environments. Centralized and flat federated systems do not efficiently coordinate large-scale, latency-sensitive and resource-constrained nodes. In this research, we present a scalable hierarchical federated Graph-Transformer architecture for effective multi-modal intrusion detection spanning UAV-edge-cloud tiers. The platform employs hierarchical aggregation among cars, UAVs, and regional edge servers for reducing communication overhead (CO) and speeding up convergence in non-IID scenarios. Meanwhile, a hybrid Graph Neural Network (GNN) and Transformer backbone is adopted to model spatial topology and temporal dynamics, and a lightweight multi-modal fusion is employed to integrate network traffic, telemetry and channel condition information. To improve the efficiency of the system, we propose adaptive aggregation scheduling and communication compression algorithms that considerably reduce bandwidth consumption and training latency. The experimental results on the CIC-IoT-2023, ToN-IoT and Edge-IIoTset datasets exhibit enhanced scalability with over 98.20% detection accuracy and up to 38.00% transmission cost reduction compared to the flat federation baselines. The work presents a scalable and system-efficient approach for next generation distributed intrusion detection in large-scale 6G vehicle ecosystems.},
}

@article {pmid42414502,
year = {2026},
author = {Sugiharto, WH and Prasetijo, AB and Susanto, H},
title = {An edge-IoT water quality index (IoT-WQI) for first-line screening: accelerating computation via deterministic mathematical equations and grouped AHP.},
journal = {Scientific reports},
volume = {},
number = {},
pages = {},
doi = {10.1038/s41598-026-61061-0},
pmid = {42414502},
issn = {2045-2322},
abstract = {Real-time water quality monitoring remains challenging due to the high latency of centralized laboratory analysis and the substantial computational payload of Machine Learning (ML) inference. While existing Internet of Things (IoT) frameworks deploy raw sensors, they frequently lack a standalone Edge calculation layer capable of operating completely independent of cloud infrastructure. This study proposes a mathematical-based Internet of Things Water Quality Index (IoT-WQI) architecture engineered specifically to act as an autonomous Fog/Edge Node for immediate first-line screening. The framework reduces continuous cloud dependency and network communication overhead by circumventing ML dependencies, replacing discrete look-up tables with mathematically continuous Gaussian and Polynomial Curve Fitting. This zero-training algorithmic optimization accelerates local computation by 83.55% over traditional linear interpolation and reduces network payload transmission bandwidth by 83.3%, while achieving an exceptional approximation fidelity ([Formula: see text]). Concurrently, a cross-sectoral Analytical Hierarchy Process (AHP) was structurally aligned with local river ecosystem profiles to extract robust contextual weights (CR = 0.02). Validated through a hardware-algorithmic stress test at the Troso River (Indonesia), the fault-tolerant topology yielded 6,779 validated records across a 6.49-hour deployment window via asynchronous queueing, achieving 98.1% of transmissions within 0-1 s end-to-end latency. Ultimately, this framework seamlessly integrates edge computational efficiency with cloud distribution frameworks, providing a highly scalable architecture for real-time anomaly triage in resource-constrained IoT environments.},
}

@article {pmid42403428,
year = {2026},
author = {Mohan Babu, A and Madhan, ES},
title = {Human digital twins in personalized and predictive healthcare: a comprehensive review of technologies, applications, and future directions.},
journal = {Frontiers in digital health},
volume = {8},
number = {},
pages = {1827007},
pmid = {42403428},
issn = {2673-253X},
abstract = {The role of real-time data, artificial intelligence, and computational modeling is discussed in this review analytics Human Digital Twins (HDTs) creation- virtual persons of personalities patients which advocate predictive simulation to forecast of physiological behavior, treatment responses, and disease tracks. A synthesis of existing knowledge is done up to the technologies is a foundation to HDTs, clinical application and implementation issues of interest to precision medicine. The conceptual basis of engineering of the digital twins is analyzed and production principles, and technologies, which allow to produce HDTs-machine. Are physiological modeling, learning and distributed cloud-based computing infrastructure identified and evaluated. Cards: cardiology, oncology, genomics and immunology are critically appraised. It is based on the comparative analysis of 35 peer-reviewed documents and technical as it was reported, HDTs have great potential in enhancing personalized prediction of side effects, optimization of clinical trial design using virtual, and scheduling of treatment cohort simulation. But, model standards, an important component of model validation, are not present interoperability, ethical governing mechanisms and regulatory avenues to clinical deployment. The main priority research directions are determined, such as the development of common-validation techniques; implementation of federated learning frameworks to support sharing of data with data privacy limitations; incorporation of multi-omics data into physiological models; and introducing open ethical review procedures. This review provides substantive evidence basis to researchers, clinicians and policy makers to market the. Knowledge about HDTs technology to population health and health care provision revolutionizes.},
}

@article {pmid42403432,
year = {2026},
author = {Sasikumar, SK and Pai, TV and Kalidasan, K and Gajendran, S},
title = {Reinforcement learning driven edge-cloud coordination for secure and energy efficient IoMT.},
journal = {Frontiers in digital health},
volume = {8},
number = {},
pages = {1824480},
pmid = {42403432},
issn = {2673-253X},
abstract = {The Internet of Medical Things (IoMT) enables sophisticated medical devices, but it also poses significant challenges in terms of data privacy, real-time processing, and energy efficiency for edge devices with limited resources. In this paper, we propose a hierarchical framework for intelligent and secure IoMT-based healthcare monitoring. At the sensor nodes, Federated Variational Mode Decomposition (VMD) is used to decompose physiological signals and locally extract high-fidelity features, ensuring data privacy. To overcome the computational limitations of microcontroller- based sensor nodes, a SparseBonsai neural network is designed for real-time classification of medical signals on the sensor nodes. A centralized orchestration layer, controlled by a Proximal Policy Optimization (PPO) reinforcement learning agent, makes dynamic decisions on whether to queue data for low-latency processing at the edge server or offload to the cloud, depending on data severity, network conditions, and battery level. To further improve energy efficiency, an advanced Sha-Dragon (Shannon-Entropy Dragonfly) optimization algorithm is proposed for resource and transmission power allocation in the IoT network. For security, a dual-layer approach is adopted: ASCON v1.2 lightweight authenticated encryption is used to secure node-to-edge communications, and a WireGuard VPN with ChaCha20-Poly1305 encryption protects data in transit to the cloud. Experimental validation on a Raspberry Pi 5 testbed with a cloud-connected laptop shows that the proposed system achieves a significant reduction in latency for critical alerts and improves the battery life of IoT nodes (8.5 days) compared to the conventional non-adaptive offloading approach. The results confirm the effectiveness of the proposed framework to facilitate energy-efficient, privacy-preserving, and real-time healthcare monitoring in IoMT.},
}

@article {pmid42404351,
year = {2026},
author = {Lan, Q and Choi, A and Ma, J and Wang, B and Zhao, Z and Jiang, X and Hsu, YC},
title = {From performance to practice: knowledge-distilled segmentator for on-premises clinical workflows.},
journal = {npj health systems},
volume = {3},
number = {},
pages = {58},
pmid = {42404351},
issn = {3005-1959},
abstract = {Deploying medical image segmentation models in routine clinical workflows is often constrained by on-premises infrastructure, where computational resources are fixed and cloud-based inference may be restricted by governance and security policies. While high-capacity models achieve strong segmentation accuracy, their computational demands hinder practical deployment and long-term maintainability in hospital environments. We present a deployment-oriented framework that leverages knowledge distillation to translate a high-performing segmentation model into a scalable family of compact student models without modifying the inference pipeline. The framework is primarily evaluated on nnU-Net, with additional validation across transformer and heterogeneous teacher-student architectures. The proposed approach preserves architectural compatibility with existing clinical systems while enabling systematic capacity reduction. We evaluate framework on a multi-site brain MRI dataset comprising 1104 3D volumes, with independent testing on 101 curated cases, and is further examined on abdominal CT to assess cross-modality generalizability. Under aggressive parameter reduction (94%), the distilled student model preserves nearly all of the teacher's segmentation accuracy (98.7%), while achieving substantial efficiency gains, including up to a 67% reduction in CPU inference latency without additional deployment overhead. These results demonstrate that knowledge distillation provides a practical and reliable pathway for converting research-grade segmentation models into maintainable, deployment-ready components for on-premises clinical workflows in real-world health systems.},
}

@article {pmid42409430,
year = {2026},
author = {Curtis, JR and Holladay, E and Mehta, T and Osborne, J},
title = {Essential Informatics Tools and Computing Infrastructure for Big Data to Advance Artificial Intelligence in Rheumatology.},
journal = {Rheumatic diseases clinics of North America},
volume = {52},
number = {3},
pages = {439-465},
doi = {10.1016/j.rdc.2026.05.001},
pmid = {42409430},
issn = {1558-3163},
abstract = {Rheumatic diseases are chronic, heterogeneous, and longitudinal, and assembling real-world evidence for effectiveness and safety for their study is best served by integrating diverse data types. This article describes the infrastructure required to support scalable, trustworthy artificial intelligence (AI) in rheumatology, emphasizing data acquisition, harmonization, linkage, privacy protection, and computational environments. We outline computing infrastructure considerations relevant to rheumatology, including hybrid on-premises and cloud architectures. Sustained progress for AI applied to rheumatology will depend on deliberate investment in shared infrastructure, longitudinal data ecosystems, and governance models that balance innovation, privacy, reproducibility, and equitable clinical value.},
}

@article {pmid42409855,
year = {2026},
author = {Bibi, M and Khan, WZ and Ul Haq, QE},
title = {Edge-assisted post-quantum authentication protocol for IoMT: a privacy-preserving and lightweight approach.},
journal = {Scientific reports},
volume = {},
number = {},
pages = {},
doi = {10.1038/s41598-026-59644-y},
pmid = {42409855},
issn = {2045-2322},
support = {NAUSS-25-R07//Naif Arab University for Security Sciences/ ; },
abstract = {The emergence and growth of the Internet of Medical Things (IoMT) have transformed the way healthcare services are delivered, but this growth comes with its own security challenges, especially with the emergence of quantum computing. This paper presents a lightweight, privacy-preserving authentication scheme for edge-based IoMT environments, integrating CRYSTALS-Kyber Key Encapsulation Mechanism (KEM) to securely establish session keys between participating entities and resilience against quantum attacks, while CRYSTALS-Dilithium is utilized as the digital signature scheme to ensure message integrity, authentication, and non-repudiation. These NIST-standardized post-quantum primitives collectively strengthen the system's resistance against quantum adversaries and enhance overall cryptographic robustness. The proposed system comprises two phases: Registration and Authentication. During registration, patients, wearable devices that utilize Physically Unclonable Functions (PUFs), and medical specialists register with a Trusted Authority via edge nodes. The authentication phase encompasses mutual authentication among patients, devices, medical specialists, edge nodes, and cloud servers, ensuring that only authorized entities access sensitive health information. In addition, the proposed scheme employs the formal security properties, such as authenticity and secrecy, are verified using the AVISPA tool under the Dolev-Yao model, with results confirming that all defined goals are achieved and no attacks are found. Performance evaluations indicate that the scheme is computationally efficient and suitable for resource-constrained IoMT devices. Furthermore, BAN logic is employed to formally verify the correctness of the authentication protocol, demonstrating that mutual authentication is securely established among the participating entities under well-defined trust assumptions. By integrating post-quantum cryptographic techniques, this authentication framework addresses emerging security challenges posed by quantum computing, aligning with the objectives of enhancing edge-based consumer electronics devices and Internet of Things (IoT) ecosystems with post-quantum cryptosystems.},
}

@article {pmid42394035,
year = {2026},
author = {Sribudda, S and Harnpicharnchai, K and Nankongnab, N and Thongkum, W},
title = {Development of a Hospital Air Quality Monitoring Application Utilizing IoT Technology and Google Workspace: A Case Study of a Hospital in Mahasarakham, Thailand.},
journal = {Studies in health technology and informatics},
volume = {338},
number = {},
pages = {409-413},
doi = {10.3233/SHTI260873},
pmid = {42394035},
issn = {1879-8365},
mesh = {*Environmental Monitoring/methods/instrumentation ; *Air Pollution, Indoor/analysis ; Thailand ; Humans ; *Internet of Things ; *Hospitals ; *Mobile Applications ; Working Conditions ; Digital Health ; },
abstract = {Hospitals are high-risk environments where Indoor Air Quality (IAQ) significantly impacts the health and safety of personnel. Inefficient ventilation and the accumulation of pollutants can lead to Sick Building Syndrome and various non-communicable diseases. Conventional manual reporting systems often limit the speed and accuracy of risk assessment. Therefore, this study aimed to develop a digital solution to provide real-time air quality data for precise occupational risk management. This action research followed the Systems Development Life Cycle (SDLC) and Database Life Cycle (DBLC) through seven phases. An IoT-based monitoring device was developed to measure eight critical parameters: CO2, CO, TVOCs, PM2.5, PM10, temperature, and relative humidity. The system utilized Google Apps Script for automated reporting and Google App Sheet for data visualization, with data stored and processed through Google Sheets and a cloud server. The developed system successfully integrated IoT sensors with Google Workspace for real-time monitoring via mobile and web interfaces. Evaluation results indicated a high level of user satisfaction among hospital personnel, with mean scores ranging from 4.03 to 4.50. The highest ratings were achieved in "System Processing Speed" and "Alignment with User Requirements" (Mean = 4.50), demonstrating that the application effectively addresses staff needs. The application proves to be an efficient tool for monitoring hospital air quality and managing occupational health risks. By integrating IoT technology with low-code platforms, the system reduces administrative workload and promotes sustainable risk management practices.},
}

*Environmental Monitoring/methods/instrumentation
*Air Pollution, Indoor/analysis
Thailand
Humans
*Internet of Things
*Hospitals
*Mobile Applications
Working Conditions
Digital Health

@article {pmid42381900,
year = {2026},
author = {Levites Strekalova, YA and Liu-Galvin, R and Khan, M and Lee, E and Alema-Mensah, E and Ofili, E},
title = {Democratizing computational skills: evaluating an asynchronous microlearning framework for cloud-based data analytics in health services research.},
journal = {Frontiers in public health},
volume = {14},
number = {},
pages = {1868973},
pmid = {42381900},
issn = {2296-2565},
support = {U24 MD015970/MD/NIMHD NIH HHS/United States ; },
mesh = {Humans ; *Health Services Research ; *Data Analytics ; *Cloud Computing ; Curriculum ; Pilot Projects ; },
abstract = {BACKGROUND: Public health is undergoing a digital transformation, with increasing reliance on data-driven decision-making that requires proficiency in computational tools. However, traditional curricula often emphasize theoretical knowledge over applied technical skills, contributing to gaps in workforce readiness. This study evaluated a pilot remote, asynchronous microlearning course designed to expand access to digital skills-specifically R and Google Colab-among students from historically underrepresented backgrounds within the Research Centers in Minority Institutions (RCMI) network.

METHODS: A three-week course, "Introduction to Cloud Data Analytics for Health Services Research," was delivered via a Learning Management System. Students (N = 19) completed three modules: (1) R and RStudio fundamentals, (2) cloud-based analysis of data from the Health Information for National Trends Survey 7 using Google Colab, and (3) scientific abstract writing. Evaluation included a pre- and post-program five-item objective knowledge assessment, retrospective self-rated competencies (1-5 scale), and a post-program satisfaction survey.

RESULTS: Mean objective knowledge scores increased significantly from 3.58 to 4.37 (p = .039). Participants reported statistically significant improvements in eight of nine self-rated competencies (p < .05), with the largest gains in installing R/RStudio and navigating the interface. Satisfaction was high across domains, particularly for "value for academic development" (M= 4.5/5.0).

CONCLUSION: Brief, asynchronous microlearning experiences can effectively build foundational computational skills and expand access to training for students in low-resourced settings. While technical competencies can be developed within short, flexible formats, more complex skills such as scientific communication may require additional instructional time and support.},
}

Humans
*Health Services Research
*Data Analytics
*Cloud Computing
Curriculum
Pilot Projects

@article {pmid42381266,
year = {2026},
author = {Chen, L and Cheng, S and Zhang, L and Li, H and Tu, Z and Cai, X},
title = {Advances in Remote Monitoring Technology Applications in Anesthesia: A Narrative Review.},
journal = {Medical science monitor : international medical journal of experimental and clinical research},
volume = {32},
number = {},
pages = {e952513},
doi = {10.12659/MSM.952513},
pmid = {42381266},
issn = {1643-3750},
mesh = {Humans ; Telemedicine ; *Anesthesia/methods ; Monitoring, Physiologic/methods ; Remote Patient Monitoring ; *Anesthesiology/methods ; Digital Health ; Internet of Things ; },
abstract = {The swift progress of digital and sensor technologies is hastening the incorporation of remote monitoring into anesthesiology. While several reviews have explored telemedicine and artificial intelligence in anesthesia, most existing summaries either focus on conceptual outlook or lack systematic comparison of technical platforms and original clinical validation data. The present review provides a comprehensive, clinically oriented synthesis of remote monitoring in anesthesia, with clear focuses on technical principles, perioperative applications, platform comparisons, and evidence-based clinical outcomes. We critically assess clinical advantages and implementation hurdles, covering the full perioperative pathway - preoperative evaluation, intraoperative observation, and postoperative recovery. A head-to-head comparison of fifth generation of cellular network technology (5G), the Internet of Things (IoT), and cloud computing platforms is presented to clarify their infrastructure demands and suitability for real-world clinical scenarios. Notably, this review summarizes available clinical evidence, including evidence from the Trial of Remote Continuous versus Intermittent National Early Warning Score Monitoring after major surgery (TRaCINg) - a feasibility randomized controlled trial whose exploratory findings suggest that continuous remote monitoring may reduce unplanned intensive care unit admissions, shorten hospital stays, and facilitate earlier detection of postoperative complications. We further propose innovative solutions including multimodal data fusion and federated learning-driven predictive analytics to overcome current limitations in data interoperability, security, and clinical effectiveness. Accordingly, this paper synthesizes the latest advances in remote monitoring technology in anesthesia, clarifies its unique value relative to existing reviews, and provides practical guidance for clinical translation and future research.},
}

Humans
Telemedicine
*Anesthesia/methods
Monitoring, Physiologic/methods
Remote Patient Monitoring
*Anesthesiology/methods
Digital Health
Internet of Things

@article {pmid42366189,
year = {2026},
author = {Arega, SY and Yan, D and Qin, T and Ali, MSI and Getahun, NK and Hassanien, MM},
title = {GEE-integrated ML classifiers evaluation for LULC change detection and CA-Markov-based future prediction in Upper Blue Nile River Basin, Ethiopia.},
journal = {Scientific reports},
volume = {},
number = {},
pages = {},
doi = {10.1038/s41598-026-58727-0},
pmid = {42366189},
issn = {2045-2322},
support = {52130907//National Science Fund Project/ ; },
abstract = {Land use and land cover (LULC) change significantly affects environmental processes and sustainable land management in river basins. The Upper Blue Nile River Basin, Ethiopia, has experienced significant LULC changes due to population growth, agricultural expansion, and deforestation. This study examines past and future LULC dynamic patterns using an integrated cloud-based framework implemented in Google Earth Engine. Classification and regression tree, random forest (RF), and support vector machine classifiers were used to process multi-temporal Landsat imagery. RF achieved the highest overall accuracy (94.4%) and kappa (0.879) in 2024. The results reveal pronounced agricultural expansion and substantial forest loss over the past two decades. Using the RF-derived LULC maps, a Cellular Automata-Markov model was calibrated and validated with strong agreement (Kappa = 0.886) to project future changes. Projections to 2034 and then 2044 indicate continued expansion of agricultural and built-up areas at the expense of forests and shrub/grasslands. The results support improved land use planning and environmental sustainability in the basin.},
}

@article {pmid42364045,
year = {2026},
author = {Lekha, PS and Sirisha, BL},
title = {Hyperspectral image processing techniques for environmental monitoring: a comprehensive review.},
journal = {Environmental monitoring and assessment},
volume = {198},
number = {7},
pages = {},
pmid = {42364045},
issn = {1573-2959},
abstract = {Remote sensing and environmental monitoring have significantly advanced with the emergence of hyperspectral image processing techniques, offering unparalleled detail in spectral analysis. Traditional remote sensing methods, such as multispectral and panchromatic imaging, often lack the spectral resolution necessary to detect subtle environmental changes. This limitation hampers the accuracy of monitoring applications such as vegetation stress, pollution detection, and land cover classification. This study reviews hyperspectral image processing techniques to enhance the accuracy of environmental monitoring. It aims to improve the detection and classification of subtle changes in land cover, vegetation health, and pollution levels. The study explores the evolving landscape of hyperspectral image processing methods and their critical role in remote sensing applications. Techniques for spectral and spatial feature extraction, dimensionality reduction, and data fusion address the complexity of hyperspectral data. Challenges like high dimensionality, noise, limited labeled data, and model interpretability are discussed. The review also highlights recent advancements, including deep learning architectures, attention mechanisms, transfer learning, generative models, and cloud-based solutions for real-time processing. Practical applications in land cover classification, vegetation health, water quality assessment, disaster response, and urban development are examined. These integrated approaches aim to enhance monitoring accuracy, efficiency, and decision-making in real-world scenarios. Future research may focus on improving real-time processing capabilities through edge computing and AI-driven models, expanding labeled datasets, and enhancing model interpretability to further advance hyperspectral imaging applications.},
}

@article {pmid42365048,
year = {2026},
author = {Mamodiya, U and Kishor, I and Jain, J and Jain, R},
title = {A machine learning-driven virtualized edge framework for real-time monitoring and anomaly detection in solar photovoltaics.},
journal = {Scientific reports},
volume = {},
number = {},
pages = {},
doi = {10.1038/s41598-026-58923-y},
pmid = {42365048},
issn = {2045-2322},
abstract = {The rapid development of solar photovoltaics requires the need to have monitoring systems because they require real time response and are also stable in a wide range of operating conditions. Typical SCADA systems are inexpensive although with limited temporal resolution and can only usually spot faults when performance has become noticeable. Machine learning solutions on clouds are more accurate but lack latency, connection reliance and are prone to privacy threats. This void creates the necessity of fine grained and real time anomaly detection. In this research, a virtualized edge architecture (VEAD) is suggested, which combines convolutional networks, the gated recurrent unit, and XGBoost classifiers into a containerized pipeline and deploys on a Raspberry Pi. The framework utilizes multi-sensor PV data to project the results into a VR-based digital twin by running the data locally. This article demonstrates that edge intelligence and immersive visualization can be used hand in hand without the need to hinder the other. The model offers quick calculations and results that are easy to decipher. This balance is checked by tests of VR-simulated faults. The detection accuracy was 96.2% and inference was under one second with an average of ~ 312 ms. There was lag in cloud baselines. Also, the edge response time was over 60% faster, which is an important difference when faults develop rapidly. Under shading, dust, and disturbances of inverter performance, performance remained constant.},
}

@article {pmid42365078,
year = {2026},
author = {Yadav, D and Sheoran, S and Aman, M and Prabuwono, AS and Khan, T},
title = {A lightweight heuristic for cost-efficient IaaS auto-scaling of small-scale web applications.},
journal = {Scientific reports},
volume = {},
number = {},
pages = {},
doi = {10.1038/s41598-026-59763-6},
pmid = {42365078},
issn = {2045-2322},
abstract = {Pay-per-use Infrastructure-as-a-Service (IaaS) makes web-application hosting affordable for small organisations, yet cost-efficient elasticity remains unsolved for deployments of two to eight virtual machine instances: enterprise auto-scalers demand weeks of traffic history and dozens of tuning parameters, while naive fixed-threshold policies react only after service degradation has begun. This paper proposes the Lightweight Adaptive Scheduling Heuristic (LASH), an O(1)-state two-phase algorithm that minimises hourly IaaS cost subject to a 200 ms P99 latency SLA. Phase 1 applies double exponential smoothing to forecast request rate one VM warm-up horizon ahead; phase 2 selects the minimum-cost instance count while a two-clause minimum-lifetime / billing-aware flag suppresses premature scale-in. LASH is evaluated against four competitive baselines (fixed-threshold, moving-average, recursive-least-squares regression, and AWS Target Tracking) in a trace-driven discrete-time simulation calibrated to AWS EC2 and Azure VM pricing, instance warm-up, and queueing behaviour, across six synthetic load profiles ([Formula: see text] seeded runs per cell; 600 simulated experiments) and, for the AWS EC2 configuration only, the real FIFA World Cup 1998 24-hour production trace ([Formula: see text] replays). In simulation, LASH dominates every baseline on cost across all six profiles and on P99 latency across all but the lowest-CoV profiles, where the regression forecaster [Formula: see text] is competitive. The mean cost reduction versus the fixed-threshold baseline is 41.9 % (BCa 95 % CI [40.7 %, 43.1 %], quantifying simulator run-to-run variability rather than deployment uncertainty), with a 23.7 % P99 latency reduction and a 75.9 % SLA-violation reduction; against a CPU-target reactive policy modelled on AWS Target Tracking the cost reduction is 13.5 %. All improvements are statistically significant under the matched-block Friedman test ([Formula: see text], Friedman [Formula: see text]) and a corroborating linear mixed-effects model on run-level data. As a simulation study, these results characterise expected behaviour under the modelling assumptions stated in the paper and are not a substitute for measurement on production infrastructure.},
}

@article {pmid42356695,
year = {2026},
author = {Banjo, O and Ghoraani, B},
title = {A Cascaded Quantized Spiking Neural Network for Real-Time ECG Arrhythmia Detection on Edge Hardware.},
journal = {Sensors (Basel, Switzerland)},
volume = {26},
number = {12},
pages = {},
doi = {10.3390/s26123723},
pmid = {42356695},
issn = {1424-8220},
mesh = {*Arrhythmias, Cardiac/diagnosis/physiopathology ; *Neural Networks, Computer ; *Electrocardiography/methods ; Humans ; Signal Processing, Computer-Assisted ; Algorithms ; },
abstract = {Wearable ECG monitors enable continuous cardiac surveillance, but most still rely on cloud-based analysis with limited on-device support for multi-class arrhythmia detection. Spiking neural networks (SNNs) are promising for low-power edge inference, yet it remains unclear how class-imbalance loss design interacts with RR-interval features in directly trained quantized SNNs, and FPGA validation in this setting is largely unexplored. We propose a quantized convolutional spiking neural network (QCSNN) for real-time arrhythmia detection on resource-constrained hardware. The model uses a dual-head architecture that jointly trains binary and four-class classifiers, subsequently reorganized into a cascaded pipeline that routes only abnormal beats to the second stage. At inference, beats classified as Normal exit at Stage 1; only beats classified as Abnormal are routed to the four-class head, so the bulk of the inference cost is absorbed by Stage 1. We evaluate two loss functions, Cross-Entropy and Focal Loss, under four RR-feature routing strategies. Without RR features, Focal Loss improves macro F1 by 2.3-2.5% over Cross-Entropy (mean Δ = +0.013 in Stage-2 macro F1; Wilcoxon two-sided p = 0.031). With RR features, this advantage largely disappears (Wilcoxon two-sided p ≥ 0.219 at all RR routings); meanwhile, RR features at the strongest routing improve Stage-2 macro F1 by +0.028 to +0.034 depending on loss function-a gain that exceeds the entire Focal-Loss-over-Cross-Entropy advantage, suggesting that RR features provide discriminative information that compensates for class imbalance at the input level. Based on clinically prioritized sensitivity, the CE:RR→Both configuration was deployed on a PYNQ-Z2 FPGA, achieving 99.02% cascaded accuracy, 11.54 ms per-beat latency, and 0.33 W accelerator power-a 31.66× power reduction and 4.01× energy reduction versus GPU inference, within 1% macro F1. These results demonstrate quantized SNNs as a practical solution for real-time edge arrhythmia monitoring that operates independently of cloud connectivity-removing the network-dependent latency, connectivity-dropout failure modes, and continuous-transmission energy burden that constrain current wearable monitors and, to our knowledge, represent one of the first systematic studies of loss-function/RR-feature interactions in directly trained SNN arrhythmia classification and one of the first FPGA deployments of a fully quantized, directly trained SNN for multi-class ECG arrhythmia detection. All code generated and used in this study has been made publicly available.},
}

*Arrhythmias, Cardiac/diagnosis/physiopathology
*Neural Networks, Computer
*Electrocardiography/methods
Humans
Signal Processing, Computer-Assisted
Algorithms

@article {pmid42356728,
year = {2026},
author = {de la Torre, JA and Rincón, F and Escolar, S and Caruso, A and Caba, J and Barba, J},
title = {Extreme Edge Computing for Secure and Private Multimodal Biometric Identification in Intelligent IoT Systems.},
journal = {Sensors (Basel, Switzerland)},
volume = {26},
number = {12},
pages = {},
doi = {10.3390/s26123756},
pmid = {42356728},
issn = {1424-8220},
support = {TSI-069100-2023-0014//European Union/ ; PID2023-148285OB-C41//Ministerio de Ciencia, Innovación y Universidades/ ; SBPLY/24/ 180225/000105//Regional Government of Castile-La Mancha/ ; GA Nº 101213323//Horizon Europe program (project MAYA)/ ; },
abstract = {The exponential growth of Internet of Things (IoT) ecosystems is driving a paradigm shift from centralized cloud computing towards decentralized architectures to mitigate latency and bandwidth constraints. While edge computing addresses some of these challenges, data transmission to local gateways still raises critical security and privacy concerns. This study explores the Compute Continuum by pushing intelligence to the extreme edge using TinyML. We propose a secure, privacy-preserving multimodal biometric authentication system designed for resource-constrained embedded devices. Our solution implements a hierarchical processing chain: an ultra-lightweight person-detection filter acts as an intelligent wake-up mechanism, followed by robust facial and voice authentication modules. Operating as a strict hierarchical pipeline, the system achieves a combined False Acceptance Rate (FAR) of just 0.12%. Experimental results on an ESP32 microcontroller demonstrate exceptional energy efficiency, requiring only 0.15 J per inference cycle. This allows the system to operate autonomously for over 39 h of continuous inference on a standard 600 mAh battery, proving the viability of standalone, privacy-by-design biometric sensors in intelligent IoT environments.},
}

@article {pmid42356775,
year = {2026},
author = {Zhang, P and Wang, R and Sun, Y and Yuan, Y},
title = {A Sensor-Aware Multi-Agent Reinforcement Learning Framework for Joint Data Offloading and Power Control in Edge-Assisted Wireless Sensor Networks.},
journal = {Sensors (Basel, Switzerland)},
volume = {26},
number = {12},
pages = {},
doi = {10.3390/s26123802},
pmid = {42356775},
issn = {1424-8220},
support = {24JCQNJC00280//Joint Funds of the Natural Science Foundation of Tianjin/ ; },
abstract = {Wireless sensor networks supported by mobile edge computing are increasingly required to process heterogeneous sensing data under stringent latency, reliability, and energy constraints. However, most existing task-offloading studies are still formulated for generic user equipment and primarily focus on uplink transmission, which is insufficient for practical sensing systems where sensor nodes continuously upload measurements while simultaneously receiving control commands, model updates, and feedback from the edge. To address this gap, this paper reformulates joint computation offloading and power control as a sensor-aware optimization problem in an edge-assisted wireless sensor network. We propose a three-layer architecture consisting of sensor nodes, access points with lightweight edge servers, and a cloud coordination layer. Each sensing task is characterized by data size, computation density, latency deadline, and sensing priority, while the optimization objective jointly minimizes long-term task delay, communication and computation energy, and packet-loss penalty under transmission power, edge resource, and residual-energy constraints. To solve the resulting mixed discrete-continuous problem, we develop a multi-agent reinforcement learning framework in which each sensor node acts as an autonomous agent and learns offloading and transmission policies with clipped proximal policy optimization, while the cloud layer performs coordinated edge-resource allocation through the alternating direction method of multipliers. In addition to delay and energy, network lifetime and sensing delivery performance are incorporated into the evaluation. Simulation results in a sensor-network monitoring scenario demonstrate that the proposed framework consistently reduces latency, lowers energy consumption, and prolongs network lifetime compared with representative baselines, highlighting its effectiveness and practical potential for intelligent sensing applications that require integrated sensing, communication, and edge computing.},
}

@article {pmid42356781,
year = {2026},
author = {Ieva, S and Loconte, D and Loseto, G and Lopomo, F and Notarnicola, M and Sblendorio, A and Scioscia, F and Ruta, M},
title = {Cloud-Edge MLOps for Diagnostic Analytics and Anomaly Detection in Smart Office Digital Twins.},
journal = {Sensors (Basel, Switzerland)},
volume = {26},
number = {12},
pages = {},
doi = {10.3390/s26123807},
pmid = {42356781},
issn = {1424-8220},
support = {NIL6S28//Regione Puglia/ ; },
abstract = {Smart buildings require intelligent and scalable solutions to monitor environmental conditions and manage increasingly complex data streams generated by distributed sensing infrastructures. In this context, the paper presents an edge-enabled Digital Twin framework for smart office environments, integrating real-time data acquisition, distributed intelligence, and machine learning-based analytics. The framework adopts a multi-layer architecture composed of a sensor layer, a cloud-edge intelligence layer, and an interaction layer, aligned with Digital Twin reference models. By enabling low-latency processing at the edge and supporting continuous model lifecycle management through Machine Learning Operations (MLOps) practices, the proposed approach overcomes key limitations of traditional cloud-centric solutions. Autoencoder-based models are deployed across the cloud-edge continuum to perform real-time anomaly detection on time-series sensor data. A prototype has been implemented in a real smart office environment, where heterogeneous environmental data are continuously collected and processed. Experimental results demonstrate effective end-to-end data flow, stable long-term operation, and reliable anomaly detection with low-latency response. The system enables real-time monitoring and data-driven analysis of environmental conditions, improving situational awareness and supporting operational decision-making. These findings confirm the effectiveness of integrating Digital Twin technologies with edge AI and MLOps principles for scalable and efficient smart building monitoring systems.},
}

@article {pmid42356796,
year = {2026},
author = {Phatcharasaksakol, B and Sittithanon, S and Pianapitham, V and Chantrapanichkul, V and Tang, J and Chancharoen, R},
title = {Design and Validation of a Cyber-Physical Medication Dispensing Platform Integrating Edge AI Verification, Distributed Control, and Cloud Synchronization.},
journal = {Sensors (Basel, Switzerland)},
volume = {26},
number = {12},
pages = {},
doi = {10.3390/s26123823},
pmid = {42356796},
issn = {1424-8220},
mesh = {Humans ; *Cloud Computing ; Intelligent Systems ; *Artificial Intelligence ; Robotics ; Medication Errors/prevention & control ; },
abstract = {Medication dispensing errors remain a significant concern in healthcare systems, particularly in elderly care and long-term medication management, where incorrect medication delivery may compromise patient safety and treatment outcomes. This study presents the design and experimental validation of a cyber-physical medication dispensing platform integrating robotic manipulation, edge AI-based visual verification, distributed motion control, and cloud synchronization. The platform combines a rotary medication storage mechanism, vacuum-based pill handling, a Klipper-based control framework, and a YOLOv8 perception subsystem deployed on a Hailo AI accelerator for real-time edge inference. Experimental evaluation was conducted under controlled laboratory conditions. Using an environment-specific validation dataset, the perception subsystem achieved a precision of 0.627, recall of 0.739, and mAP@0.5 of 0.786. An adaptive verification strategy was subsequently evaluated to improve dispensing verification under varying pill occupancy conditions. End-to-end system testing comprising 80 dispensing trials achieved an overall dispensing success rate of 86.25%, with no incorrect dispensing events observed. The results demonstrate the feasibility of integrating edge AI verification, distributed control, and cloud connectivity within a cyber-physical medication dispensing platform. The presented system provides a foundation for future research on perception-assisted medication dispensing, long-term deployment, and clinical validation in smart healthcare environments.},
}

Humans
*Cloud Computing
Intelligent Systems
*Artificial Intelligence
Robotics
Medication Errors/prevention & control

@article {pmid42356908,
year = {2026},
author = {Xue, J and Huang, Y and Guo, Y and Lin, Z and Tian, B},
title = {Dual-Time-Scale Cloud-Edge-End Collaborative Task Offloading for Multi-AGV Intelligent Warehousing in Industrial Internet of Things.},
journal = {Sensors (Basel, Switzerland)},
volume = {26},
number = {12},
pages = {},
doi = {10.3390/s26123936},
pmid = {42356908},
issn = {1424-8220},
support = {62471140//National Natural Science Foundation of China/ ; 2024J01250//Natural Science Foundation of Fujian Province/ ; },
abstract = {In embodied-intelligence Industrial Internet of Things (IIoT), multi-AGV intelligent warehousing requires continuous processing of latency-sensitive tasks, such as environmental perception, inventory monitoring, and anomaly detection. Due to limited onboard computing capability and energy capacity, purely local execution can hardly satisfy real-time requirements, whereas fully cloud-based processing may incur excessive transmission delay and backhaul overhead. To address this issue, this paper investigates the joint optimization of AGV service-point migration and task offloading under a cloud-edge-end collaborative architecture. Considering the impact of service-point selection on wireless access, MEC resources, movement delay, and energy consumption, as well as the effect of offloading decisions on transmission, computation, and AGV-side energy cost, a dual-time-scale optimization model is formulated to minimize the long-term accumulated system delay while satisfying task latency and AGV energy constraints. To solve the resulting mixed discrete problem, a DPSO-MAPPO algorithm is proposed, where DPSO searches service-point plans satisfying movement and conflict constraints at the slow time scale, and MAPPO learns coordinated multi-AGV offloading policies at the fast time scale. The delay and energy feedback further enables coordination between the two types of decisions. Simulation results show that the proposed algorithm converges stably, reduces system delay by 13.55% compared with benchmark algorithms, and improves total energy consumption and energy-violation control.},
}

@article {pmid42359896,
year = {2026},
author = {Ďuráčiová, R and Capandová, M and Berka, K and Svobodová, R and Slanináková, T and Kováč, K and Antol, M and Hejtmánek, L},
title = {Foldify: Web Application for Protein Structure Prediction.},
journal = {Journal of chemical information and modeling},
volume = {},
number = {},
pages = {},
doi = {10.1021/acs.jcim.6c01154},
pmid = {42359896},
issn = {1549-960X},
abstract = {Protein structure prediction models released in recent years have presented tectonic changes in the field of structural biology. However, their potential has not yet been harnessed to its fullest due to their demands on hardware and technical expertise required for their usage. In this paper, we present Foldify, which makes prediction models accessible, integrating AlphaFold 3, AlphaFold 2, ColabFold, OmegaFold, and ESMFold into a single user-friendly, easy-to-use graphical interface, and ensures their stable operation within a scalable high-performance computing environment. Foldify accepts protein sequences, submitted through a web-based graphical interface as input, and allows executing multiple prediction models on the same protein sequence. The predicted protein structures can be directly visualized online through Mol* Viewer or can be downloaded from the website. Furthermore, the multiresult comparison mode allows visualization of multiple predicted structures in a single Mol* window, accompanied by qualitative metrics of the models' prediction similarity. The Foldify application is freely available at https://foldify-open.cloud.e-infra.cz/ with no login required.},
}

@article {pmid42360507,
year = {2026},
author = {Fernandes, JEA and Nagata, T and Melo, FL},
title = {Mining viruses in public databases unveils the diversity within the Deltaflexiviridae family.},
journal = {Archives of virology},
volume = {171},
number = {7},
pages = {},
pmid = {42360507},
issn = {1432-8798},
mesh = {Phylogeny ; *Genome, Viral ; *Genetic Variation ; Data Mining ; *Flexiviridae/genetics/classification/isolation & purification ; Databases, Genetic ; Evolution, Molecular ; },
abstract = {Cloud computing platforms aided the scalability and applicability of viral mining in genomic databases. The Serratus project reported SRA accessions that may contain viral sequences. This study analyzed SRA accessions containing sequences similar to Tymovirales members to verify the presence of sequence data derived from viral genomes. All steps in the genome mining analysis were performed by a pipeline running on virtual machines hosted on the Google Cloud Platform. Manual curation of the pipeline output discovered 111 putative genomes in the analyzed SRAs. Among the genomes identified, four were classified as isolates within the Betaflexiviridae family, and two were putative new members of the Alphaflexiviridae family. Another four sequences were likely classified in a family not yet accepted by the ICTV. The phylogenetic reconstruction of the Deltaflexiviridae family revealed three distinct clades, one of them containing 81 genomes (34 putative new species), a second clade with 18 novel genomes (7 putative new species), and a third with one putative new species. Given the high divergence between these three groups, we suggest the establishment of a new family, "Epsilonflexiviridae", and the split of the Deltaflexiviridae family into two by establishing the family "Zetaflexiviridae". The results of this work provide insight into important aspects of the evolutionary history of the order Tymovirales and offer new ways for virus-mining projects in genomic databases.},
}

Phylogeny
*Genome, Viral
*Genetic Variation
Data Mining
*Flexiviridae/genetics/classification/isolation & purification
Databases, Genetic
Evolution, Molecular

@article {pmid42345944,
year = {2026},
author = {Kong, HJ},
title = {Mining the Public Mind: A Text-Mining Approach to Dental Implants and Dentures.},
journal = {Dentistry journal},
volume = {14},
number = {6},
pages = {},
pmid = {42345944},
issn = {2304-6767},
support = {2025//This research was supported by Development Fund Program of Wonkwang University College of Dentistry funded by Dentium Co., Ltd./ ; },
abstract = {Background/Objectives: This study aimed to comparatively analyze online information regarding dental implants and dentures utilizing text-mining techniques. Methods: An automated text-mining program was employed to collect and process data using the Korean keywords for "implant" and "denture." Data sources included major search engines, social networking services, and YouTube (Google LLC, Mountain View, CA, USA). A total of 9941 data points for dental implants and 9783 for dentures were retrieved. The analytical approach included word cloud generation, term frequency-inverse document frequency (TF-IDF) analysis, semantic network analysis, and sentiment analysis. Results: For implants, "dental clinic," "treatment," "surgery," and "insurance" emerged as highly relevant keywords. In contrast, queries regarding dentures frequently included the term "implant," alongside top-ranking, age-related terms such as "abnormality" and "discomfort." TF-IDF analysis revealed that "surgery" and "procedure" ranked higher for implants, whereas "insurance" ranked higher for dentures. Sentiment analysis indicated a predominantly positive public perception of implants (63.09% positive, 36.91% negative), whereas dentures elicited a largely negative sentiment (40.70% positive, 59.30% negative). Conclusions: The text-mining analysis revealed distinct public perceptions regarding the two treatments. Implants were primarily associated with surgical procedures and positive sentiments, whereas dentures were more closely linked to insurance considerations and negative experiences.},
}

@article {pmid42348546,
year = {2026},
author = {C M, V and R P, AK and D, P and P S, P},
title = {Multiobjective dynamic resource allocation in cloud computing using Harris Hawk Optimization Algorithm (MDLB-HHO).},
journal = {PloS one},
volume = {21},
number = {6},
pages = {e0351653},
doi = {10.1371/journal.pone.0351653},
pmid = {42348546},
issn = {1932-6203},
mesh = {*Cloud Computing ; *Algorithms ; *Resource Allocation/methods ; Soft Computing ; Workload ; },
abstract = {To increase cloud computing utilization and performance, efficient load balancing and resource distribution techniques are essential. Dynamic load balancing and resource allocation in cloud systems is necessary due to a number of reasons, but this is not an easy and straightforward task. The primary goal of dynamic load balancing of cloud systems is to optimize the workload and resource utilization. The Harris Hawks Optimization (HHO) algorithm is a dynamic method of allocating the workloads to the virtual machines (VMs) according to the workload distribution and the use of the resources. The comparison of experimental analysis and other load-balancing methods shows that the HHO algorithm can be used to control dynamic load balancing in a rather efficient and effective way. With such technical developments, there has been a decrease in time taken to respond as well as the use of resources. The suggested solution is a cost-effective and efficient solution to the load-balancing problem in dynamic conditions and is based on the collaborative hawks hunting behavior. The system converts the resource allocation scheme to the changeable cloud application requirements. This is achieved by a multiobjective fitness function which aims to maximize the efficiency of resources, minimize the response time and resource usage. The primary objective of the study is to ensure that the clouds services become effective and sustainable. The Harris Hawks discover the most optimal distribution techniques of activities by closely observing the space of solutions. They then apply positional updates and iterative interactions to adapt to changing workloads. The system dynamically assigns jobs to virtual machines (VMs) without compromising load balance and efficient resource use through the use of the cooperative search behavior of the hawks. The proposed solution effectively manages the cases when the task requirements are constantly changing. Applying a multiobjective fitness function greatly improves key performance metrics like overall performance, resource usage, and reaction time. This study demonstrates how the HHO algorithm increases the effectiveness and robustness of cloud-based services in dynamic operational environments.},
}

*Cloud Computing
*Algorithms
*Resource Allocation/methods
Soft Computing
Workload

@article {pmid42350506,
year = {2026},
author = {Chen, R},
title = {EdgeFusionNet: real-time multimodal feedback for table tennis training via lightweight cross-modal attention fusion on edge-cloud collaborative architecture.},
journal = {Scientific reports},
volume = {},
number = {},
pages = {},
doi = {10.1038/s41598-026-58591-y},
pmid = {42350506},
issn = {2045-2322},
abstract = {Real-time biomechanical feedback during table tennis training demands both low latency and high recognition accuracy, yet existing systems sacrifice one for the other due to cloud-transmission delays and the computational constraints of edge devices. This paper presents EdgeFusionNet, an integrated edge-cloud collaborative architecture that delivers actionable stroke-level feedback within 32 ms under realistic network conditions. At its core, a lightweight cross-modal attention fusion network (LCA-FNet) fuses temporally aligned features from high-speed vision, inertial measurement, and surface electromyography streams through shared-projection cross-modal attention and adaptive channel gating, achieving 93.6 ± 0.4% recognition accuracy (Macro-F1 = 0.927 ± 0.005, mean ± SD over five seeds) across seven canonical stroke types with only 1.48 million parameters, and retaining 90.4 ± 2.3% accuracy under a strict leave-one-subject-out evaluation. A hardware-triggered synchronization mechanism maintains sub-millisecond cross-modal alignment, while a two-level knowledge distillation strategy recovers 97.8% of the cloud-resident teacher model's accuracy after aggressive structural compression. An adaptive computation offloading agent, trained via Q-learning with explicitly defined state, action and reward spaces, dynamically partitions inference between the edge node and cloud server based on prediction entropy and network quality, sustaining sub-32 ms P95 end-to-end latency even at 5 Mbps uplink bandwidth. Field deployment over thirty training sessions with twelve athletes per study arm confirmed ecological validity, yielding 91.8 ± 0.7% accuracy under uncontrolled gymnasium conditions, a Cohen's kappa of 0.874 against the consensus of two expert coaches (whose own inter-coach kappa was 0.892), and a 10.3-percentage-point gain in standardized multi-ball hit rate over a matched control group after four weeks (p < 0.001). These results demonstrate that principled co-design of multimodal fusion, model compression, and adaptive offloading can bridge the gap between laboratory-grade recognition performance and the stringent latency requirements of live athletic training.},
}

@article {pmid42350521,
year = {2026},
author = {Ambika, B and Pandian, SMV and Sundaravadivel, P and Vinothkumar, ES},
title = {Federated deep reinforcement learning enabled hierarchical Edge-Fog-Cloud architecture for intelligent task offloading in 6G networks.},
journal = {Scientific reports},
volume = {},
number = {},
pages = {},
doi = {10.1038/s41598-026-57660-6},
pmid = {42350521},
issn = {2045-2322},
abstract = {Latency sensitive, computation intensive and mobility aware applications in Edge Fog Cloud environments have increased the demand to develop intelligent offloading task mechanism that can dynamically scale to changing network conditions whilst remaining scalable, energy efficient, and preserving data privacy. Traditional, heuristic and centralized based learning offloading methods frequently have problems in accommodating heterogeneous workloads, non- stationary environments and privacy limitations associated with the next generation distributed computer system. In order to overcome these drawbacks, the present paper will suggest a Federated Deep Q-Learning (FDQL)-based task offloading framework, which incorporates deep reinforcement learning and federated learning to support adaptive, decentralized and privacy-conscious decision-making across hierarchical Edge Fog Cloud architectures. The framework proposed solves task offloading as a Markov Decision Process, with the execution decisions being trained based on the joint consideration of the latency, bandwidth availability, queue length, computational load, and energy state, as well as user mobility, without sharing raw data during federated model aggregation. In comparison to the current CNN-, LSTM-, SVM-, and rule-based methods, which use fixed threshold values or rely on centralized training, the FDQL architecture allows collaborative learning between distributed edge nodes, enhancing generalization as well as resilience as network conditions evolve. Large-scale experimental analysis is performed using a trace-driven simulation based on a publicly available task offloading dataset of tasks and the performance is evaluated based on the latency, energy consumption, task success rate, robustness analysis, and computational efficiency. Experimental findings indicate that the proposed FDQL framework demonstrates improved performance under distributed and resource-constrained environments compared to baseline approaches since shorter latency, increased energy efficiency, and more predictable execution-layer selection are achieved. The significance of federated learning, mobility awareness, and bandwidth-aware optimization in the stability of the performance is also confirmed by ablation studies. In order to achieve a better level of transparency and trustworthiness, SLA-based confusion matrix analysis and ROC analysis are performed as well as SHAP-based explainability analysis, which proves that the decisions made by FDQL are based on physically interesting, as well as SLA-relevant features, like latency, bandwidth, and resource use. All in all, the designed FDQL framework is a successful, interpretable, and scalable approach to intelligent task offloading, so it would fit perfectly into the implementation of the 6G-enabled application, such as smart cities, industrial internet of things, and autonomous systems in the future.},
}

@article {pmid42342783,
year = {2026},
author = {Kalik, EM and Izadkhah, H and Karimpour, J},
title = {Multi-objective task scheduling using SBA-based deep reinforcement learning in cloud computing.},
journal = {Scientific reports},
volume = {},
number = {},
pages = {},
doi = {10.1038/s41598-026-58601-z},
pmid = {42342783},
issn = {2045-2322},
abstract = {Cloud computing is a key enabler of modern computing services, offering scalability and flexibility. However, efficient management of cloud resources remains challenging due to limited capacity and the increasing number of tasks requiring timely execution. An effective task scheduling strategy is therefore essential to improve resource allocation and utilization, reduce operational costs and energy consumption, and support high availability-especially for long-term jobs. In this paper, we propose a new scheduling approach that combines a Social-Based Algorithm (SBA) with Deep Reinforcement Learning (DRL), referred to as SBA-DRL. This method allocates tasks to resources by learning from workload patterns and adapting to workload characteristics in a batch scheduling context. We evaluate SBA-DRL using both a synthetic dataset and the real-world Google Cloud Jobs (GoCJ) under workloads ranging from 200 to 1,000 tasks. On the synthetic dataset, our method reduces cost by 20.21% and energy consumption by 25.31%, while improving resource utilization by 9.36%. On the GoCJ dataset, it achieves up to 28.94% lower cost, 8.16% less energy use, and a 14.04% increase in resource utilization. In both cases, SBA-DRL also demonstrates better performance in resource allocation and high-availability management compared to existing heuristics, meta-heuristics, hybrid, and machine learning-based schedulers. These results indicate that the proposed SBA-DRL approach effectively addresses key challenges in cloud task scheduling, offering a practical solution to enhance the efficiency and sustainability of cloud systems.},
}

@article {pmid42335364,
year = {2026},
author = {Cranford, HM and Pietka, T and de Wilde, L and Lawrence, M and Ekpo, LL and Ellis, EM},
title = {US Virgin Islands Launches Modernized NBS 7 Disease Surveillance System to Transform Public Health: Implementation Report.},
journal = {JMIR medical informatics},
volume = {14},
number = {},
pages = {e85365},
doi = {10.2196/85365},
pmid = {42335364},
issn = {2291-9694},
mesh = {Humans ; United States Virgin Islands ; *Public Health/methods ; Public Health Infrastructure ; *Population Surveillance/methods ; },
abstract = {BACKGROUND: During January 2024, the US Virgin Islands (USVI) Department of Health (VIDOH) identified a critical need to maintain the cloud-hosted National Electronic Disease Surveillance System Base System (NBS) instance and support the local data modernization initiative. After consulting with federal partners and subject matter experts, VIDOH's leadership chose to migrate the integrated disease surveillance system to a new platform hosted on Amazon Web Services (AWS) and update the NBS instance to the most advanced version, NBS 7.

OBJECTIVE: The primary aim was to support a USVI disease surveillance system that is modern, functional, and cost-efficient by migrating the VIDOH NBS instance from a vendor-managed environment to a jurisdiction-managed AWS cloud-based infrastructure while upgrading to NBS 7.

METHODS: The VIDOH implemented a phased migration strategy that included planning and cost-benefit assessment, deployment of NBS 7 within AWS, database migration, validation and optimization, and staged reonboarding of electronic reporting facilities.

UNLABELLED: The USVI NBS 7 instance went live on May 6, 2025, with USVI becoming the first US jurisdiction using AWS for implementation of NBS 7 and the second using NBS 7 in production, overall. Benefits of this change included nearly 90% cost savings (preliminarily estimated at 80%), additional bandwidth, real-time data ingestion and updates, an opportunity to build local informatics capacity, and the ability to have greater autonomy over the data and its end points. To date, the VIDOH successfully reonboarded 106 of 109 (97%) previously connected electronic reporting facilities and onboarded 1 new reporting laboratory previously unable to connect due to interoperability barriers.

CONCLUSIONS: Updating the USVI database to NBS 7 in a locally owned, cloud-hosted, AWS environment has improved disease surveillance specifically by providing the most up-to-date Centers for Disease Control and Prevention-supported data system, improving timeliness of reporting by offering local providers more flexibility in electronic reporting options, and giving USVI direct control over workflow decision functionality. Furthermore, improved interoperability and maintaining a cloud-based platform were additional benefits of the database migration. This important investment in public health infrastructure will allow USVI public health professionals, clinicians, policymakers, and other stakeholders to be able to monitor and respond to disease threats quickly and inform appropriate public health action.},
}

Humans
United States Virgin Islands
*Public Health/methods
Public Health Infrastructure
*Population Surveillance/methods

@article {pmid42337308,
year = {2026},
author = {Sachin, R and Jagawat, RS and Koganti, A and Reddy, VV and Sandhya, S and Abhishek, S},
title = {Thermo-CR: real-time physics-based cloud shadow removal via thermodynamic atmospheric modelling and multi-source fusion.},
journal = {Scientific reports},
volume = {},
number = {},
pages = {},
doi = {10.1038/s41598-026-58833-z},
pmid = {42337308},
issn = {2045-2322},
abstract = {Spaceborne optical sensors provide continuous Earth observation, but atmospheric interference still limits their practical reliability. On average, clouds cover 67% of the Earth's surface. This constant coverage degrades the data continuity needed for precision agriculture, disaster monitoring, and proactive Internet of Things (IoT) systems. Recent deep generative networks produce visually appealing cloud-free images. However, when faced with thick clouds ([Formula: see text] opacity), these models often hallucinate topologies. They synthesize statistical guesses instead of recovering the actual ground reflectance. For high-stakes telemetry, predictable failure is safer than an undetected hallucination. This paper introduces Thermo-Cloud Removal (Thermo-CR), a real-time cloud removal framework. It integrates Radiative Transfer inversion, weather-driven transmission estimates, geographic priors, and multi-scale fusion to restore optical imagery without requiring Synthetic Aperture Radar (SAR). Thermo-CR treats the cloudy atmosphere as a thermodynamic medium. By pulling live meteorological telemetry (Relative Humidity (RH) and Temperature (T)) through the Open-Meteo REST API, the system calculates optical depth and performs a deterministic inversion of the Radiative Transfer Model. Pure inverse models amplify noise under extreme occlusion ([Formula: see text]). To prevent this, we apply a Global Positioning System (GPS)-anchored multi-scale fusion with clear-sky temporal priors. We evaluated Thermo-CR on a synthetically occluded paired dataset covering varied topologies (Amazon, London, Seattle). The system degrades predictably under 90% cloud cover and avoids structural hallucination. It achieves an average Structural Similarity Index Measure (SSIM) of 0.9925 and a Peak Signal-to-Noise Ratio (PSNR) of 55.94 dB in under 13 milliseconds per frame, outperforming standard Dark Channel baselines.},
}

@article {pmid42337358,
year = {2026},
author = {Mani, RT and Palimar, V and Singh, S and Dazzle, AJ and Kanmani, RS and Pai, MS and Shwetha, TS and Krishnan, MN and Vinjamuri, R and Satyen, L},
title = {A cloud-based two-layer text classification framework for mental health screening with sarcasm and emoji-aware sentiment analysis.},
journal = {Scientific reports},
volume = {},
number = {},
pages = {},
doi = {10.1038/s41598-026-58453-7},
pmid = {42337358},
issn = {2045-2322},
abstract = {The increasing use of digital communication platforms has led individuals to express emotions and mental health concerns through text containing implicit emotional cues, informal language, and non-standard expressions. Traditional sentiment analysis systems often struggle to capture these contextual nuances, limiting their effectiveness in mental health-related text analysis . To address this challenge, this study proposes a two-layer framework that combines Azure Sentiment Analysis and Azure Custom Text Classification for sentiment and mental health-related text categorization. In the first layer, user-generated text is classified into positive, neutral, or negative sentiment categories using Azure Sentiment Analysis. Text identified as negative is subsequently analysed using Azure Custom Text Classification to categorize content into predefined mental health-related classes, including Anxiety, Depression, PTSD, Social Anxiety Disorder, and Suicidal Ideation and Behaviour. The proposed framework aims to provide a structured approach for identifying linguistic patterns associated with mental health-related discussions and supporting mental health screening and triage applications. Experimental evaluation using an 80% training and 20% testing split achieved an overall Precision, Recall, and F1-score of 96.97%. Class-level evaluation demonstrated strong performance across multiple categories, with F1-scores ranging from 0.94 to 1.000. The findings indicate that the proposed architecture can effectively classify mental health-related textual content within the evaluated dataset while providing a scalable framework for automated sentiment and text classification. The study contributes to the growing field of intelligent emotional computing and highlights the potential of cloud-based natural language processing tools for mental health-related text analytics . The reported results are limited to the evaluated dataset and should be interpreted as a text classification and screening approach rather than a clinical diagnostic system. This manuscript presents the computational component of a broader mixed-methods study registered under CTRI/2024/06/068766, titled "Exploring Mental Health Status in a Selected Population: A Corpus Analysis Combining Forensic Linguistics and Psychology - a Mixed Method Study." The current work focuses on the development and validation of an AI-based diagnostic tool for mental health assessment using synthetic and anonymized textual data, constituting a secondary objective of the registered protocol. Registry: Clinical Trials Registry- India (CTRI) Trial Registration Number: CTRI/2024/06/068766 Date of Registration: 12.06.2024.},
}

@article {pmid42326369,
year = {2026},
author = {Morales-Guerra, J and Soto-Perdomo, J and Botero-Valencia, J and Reyes-Vera, E and Pearce, JM},
title = {Low-cost embedded system for spectral power distribution reconstruction for controlled environmental agriculture using a multispectral sensor and cloud-based deep learning.},
journal = {HardwareX},
volume = {26},
number = {},
pages = {e00786},
pmid = {42326369},
issn = {2468-0672},
abstract = {This work presents an open-source device for acquiring, correcting, and reconstructing the spectral power distribution (SPD) of LED sources used in controlled environmental agriculture. Unlike direct measurement spectrometers, the system employs a low-cost multispectral sensor (AS7265x, 18 channels, 410-940 nm) to acquire sparse band-integrated data, which are subsequently processed through a two-stage machine learning pipeline to infer a dense SPD representation. The sensor is integrated into an embedded platform that performs spectral acquisition, processing, wireless transmission, and remote visualization. Comparison with a reference spectrometer revealed non-linearities and some minor limits to the agreement between sensor data and ground-truth spectra. To address this, a correction stage based on a multilayer perceptron (MLP) implemented with TensorFlow Lite Micro was developed, reducing the RMSE from 0.183 to 0.035 and improving the reliability of the data. Complementary environmental monitoring was included using a BME688 sensor to record temperature, humidity, and gas concentration, serving as a reference to detect and correlate anomalies in SPD measurements under extreme environmental conditions. All data were transmitted to a back-end server for processing. Spectral reconstruction was performed in the cloud using a one-dimensional convolutional neural network (1D-CNN) trained on horticultural LED spectra and physically inspired synthetic spectra representative of CEA. The model achieved an RMSE of 0.0135, confirming high precision within the target application domain and demonstrating a scalable and cost-effective solution for spectral monitoring in controlled agricultural environments.},
}

@article {pmid42327623,
year = {2026},
author = {Cordier, BA and Benton, ES and Dysinger, EJ and McWeeney, SK and , },
title = {Data-derived Identity Verification as a principle for the dissemination, harmonization, and artificial intelligence reuse of sensitive biomedical data.},
journal = {JAMIA open},
volume = {9},
number = {3},
pages = {ooag087},
pmid = {42327623},
issn = {2574-2531},
abstract = {OBJECTIVES: To facilitate responsible biomedical data sharing, large-scale data harmonization, and cutting-edge artificial intelligence (AI) applications by augmenting controlled-access data dissemination models (DDMs) with Data-derived Identity Verification (DIVe).

MATERIALS AND METHODS: We developed a watermarking-based DIVe system that embeds a cryptographically verifiable and traceable identity within each data file. This system was deployed for the Artificial Intelligence Ready and Exploratory Atlas for Diabetes Insights Consortium and currently watermarks a 3.8TB multimodal dataset for each approved data request.

RESULTS: Over 1.8 petabytes, spanning over 900 data copies and 160 million files, have been uniquely watermarked and disseminated. Private provenance tracing, a novel data sharing license, and educational materials extend trust and accountability beyond controlled environments.

DISCUSSION: Typical controlled-access DDMs assume data remain at rest. Data-derived Identity Verification provides an organizing principle for augmenting DDMs such that data can move while maintaining protection from data leaks.

CONCLUSION: Watermarking-based DIVe supports secure, traceable dissemination of sensitive data, supporting the development of AI technologies in biomedicine.},
}

@article {pmid42328542,
year = {2026},
author = {Quan, X and Pan, J},
title = {Enhancing healthcare quality and management: AI-driven smart consortium practices at Dapeng Medical Group, Shenzhen.},
journal = {Frontiers in medicine},
volume = {13},
number = {},
pages = {1680509},
pmid = {42328542},
issn = {2296-858X},
abstract = {OBJECTIVE: To evaluate the impact of smart healthcare consortium management practices, utilizing artificial intelligence (AI) to enhance healthcare service quality and operational efficiency at Dapeng Medical Group, Shenzhen.

DESIGN: A case study analysis of smart healthcare initiatives focusing on AI-driven improvements in patient care and management processes.

SITE: Dapeng Medical Group, a comprehensive medical institution in Shenzhen, China.

PARTICIPANTS: Patients receiving care within the Dapeng Medical Group, along with healthcare providers and administrative staff involved in the implementation of the smart healthcare consortium.

METHODS: The initiative incorporated advanced technologies, including AI, cloud computing, and the Internet of Things (IoT), to streamline healthcare operations. Key performance metrics such as patient satisfaction, waiting times, diagnostic accuracy, and resource utilization were monitored to assess the effectiveness of the smart healthcare consortium. To clarify the effectiveness of Dapeng Medical Group's artificial intelligence technologies in smart healthcare consortium management, this study compares relevant management data from before AI adoption (2022) and after AI enablement (2025).

RESULTS: The implementation led to a 25% increase in patient satisfaction, 14% reduction in average waiting times and 34% reduction in the consultation duration, a 29% reduction in error rate of online pre-diagnostic precision.

CONCLUSION: The integration of AI and smart technologies into healthcare management significantly improved service quality, operational efficiency, and patient outcomes at Dapeng Medical Group. These findings suggest that smart healthcare consortia can serve as a model for future healthcare innovations, ensuring high-quality, efficient, and patient-centered care.},
}

@article {pmid42330031,
year = {2026},
author = {Luo, J and Xie, J},
title = {Utilizing Adaptive Machine Learning Algorithms for Information Risk Warning and Network Security Scenario Awareness in Cloud Computing Environments.},
journal = {Journal of visualized experiments : JoVE},
volume = {},
number = {232},
pages = {},
doi = {10.3791/69633},
pmid = {42330031},
issn = {1940-087X},
mesh = {*Cloud Computing ; *Machine Learning ; *Computer Security ; Adaptive Algorithms ; Soft Computing ; Classification Algorithms ; Algorithms ; },
abstract = {This study proposes a novel framework for network security situational awareness and risk warning in cloud computing environments, integrating adaptive Machine Learning (ML), Hierarchical Multi-Label Classification (HMC), and a dynamic trust evaluation mechanism based on the cloud model. The complexity, diversity, and real-time nature of emerging cyberattacks-such as zero-day exploits, distributed denial-of-service (DDoS), and botnets-pose significant challenges to traditional rule-based and static detection methods. To address these challenges, we developed an effective SDN-based cloud architecture utilizing the Ryu OpenFlow controller and OpenFlow switches. This architecture enables real-time link information collection, dynamic scheduling, and scalable, reliable data transmission. The hierarchical classification framework suggested can break multiclass problems into binary tasks, alleviating the effect of sample imbalance and enhancing the recognition of low-frequency attacks, including User to Root (U2R). Ensemble learning techniques, including AdaBoost and Bagging, further enhance detection accuracy for fine-grained attack types. Experiments conducted on DDoS datasets, cloud traffic data, and simulations in Mininet and EstiNet demonstrate that the combined ML-HMC-trust approach significantly improves detection precision, reduces false positives, and enables real-time response. These results confirm that integrating adaptive learning, hierarchical classification, and dynamic trust evaluation provides a robust and scalable solution for securing large-scale cloud platforms.},
}

*Cloud Computing
*Machine Learning
*Computer Security
Adaptive Algorithms
Soft Computing
Classification Algorithms
Algorithms

@article {pmid42322262,
year = {2026},
author = {Qu, H and Duan, J and Wang, J and Huang, Y and Tang, J and Huang, Z and Li, R and Zhou, C and Feng, C and Tao, L and Zhang, J and Ma, G and Wan, H and Yang, D and Wang, H},
title = {Steady, Flexible Memristor with Lead-Free Perovskite for Bionic Perception.},
journal = {ACS applied materials & interfaces},
volume = {},
number = {},
pages = {},
doi = {10.1021/acsami.6c01730},
pmid = {42322262},
issn = {1944-8252},
abstract = {The advancement of artificial intelligence and cloud computing aspires multifunctional components to face complex problems and bionic perception. As one of the emerging neural synapse devices, the perovskite memristor has great potential in information storage and brain-like learning due to its merits, such as low energy consumption and quick processing. Herein, the PET/Graphene/Cs2AgBiBr6/Ag flexible memristor has been successfully fabricated via low-temperature processing (∼100 °C), exhibiting pronounced resistive switching characteristics (ON/OFF ratio: ≈10[2]), long-term environmental stability (≥8 months), and excellent mechanical bending performance (600 bending cycles). The synaptic plasticity in the device is also verified, including paired pulse facilitation (PPF) and spike-timing-dependent plasticity (STDP). By applying resistance memory characteristics, the memristor achieves fear emulation, enabling the realization of key characteristics including "self-extinction", "generalization", and "avoidance". Finally, sound perception is simulated via imposing different signals controlled by different voltage parameters (frequency and amplitude) and light illuminance. Sound characteristics including "pitch", "loudness" and "timbre" are successfully determined. Our work proposes a new application avenue for lead-free flexible perovskite memristors in multimodal intelligent perception and biological simulation.},
}

@article {pmid42323378,
year = {2026},
author = {Albugmi, A},
title = {Learning-based orchestration for low-latency AI deployment in hybrid cloud-edge platforms.},
journal = {Scientific reports},
volume = {},
number = {},
pages = {},
doi = {10.1038/s41598-026-58531-w},
pmid = {42323378},
issn = {2045-2322},
abstract = {The rapid growth of AI-driven applications in hybrid cloud-edge environments poses substantial challenges to ensuring low latency, high throughput, and effective resource utilization. Conventional deployment models, which are typically fixed or policy-driven, are not sufficiently flexible to respond dynamically to changing workloads and heterogeneous hardware environments. In this work, we introduce and analyze a resource-conscious deep learning-based scheduling system for managing the deployment of AI models on distributed cloud edges. The framework improves inference performance by leveraging real-time system telemetry and model features generated by benchmarks, while maintaining quality of service (QoS) compliance. The proposed system uses a fully connected neural network trained on structured features derived from the MLPerf Inference Benchmark, including compute complexity, memory footprint, and input dimensions. It is guided by real-time data from a hybrid infrastructure (NVIDIA A100/V100 GPUs and Jetson Xavier edge devices) to inform scheduling. Four MLPerf inference workloads - ResNet 50, BERT, SSD ResNet34, and DLRM - were tested and contrasted across various batch sizes and latency thresholds. Generalization experiments with unseen models such as GPT 2 and YOLOv5 yielded > 90% success rates in deployment, with the latency reduction and throughput gain results as presented above. Results of the generalization experiments with unseen models, including GPT 2 and YOLOv5, demonstrated deployment success rates > 90% for the various profiling conditions evaluated, with the latency reduction and throughput improvements as shown above. The results show that learning-based orchestration can be used to deliver space- and resource-aware orchestration solutions that are adaptive for low-latency deployment of AI services in hybrid cloud edge systems, but the effectiveness of the solution will depend on the representativeness of the profiling data and similarity of training and deployment environments.},
}

@article {pmid42317630,
year = {2026},
author = {Ha, T and Nketia, KA and Neudorf, S and Shirtliffe, SJ},
title = {Cropland mask dataset for the Canadian Prairies derived from Google satellite embedding imagery.},
journal = {Data in brief},
volume = {67},
number = {},
pages = {112946},
pmid = {42317630},
issn = {2352-3409},
abstract = {This article presents a spatially explicit persistent cropland mask for the Canadian prairies covering Alberta, Saskatchewan, Manitoba. Data was generated using 64-band embedding data from AlphaEarth with Agri-Food Canada (AAFC) Annual Crop Inventory used for label generation. Stratified-random points from across the prairies were used with a Random Forest classifier to determine cropland and noncropland areas. The trained models were applied across the prairies from 2017 to 2024 in an annual wall-to-wall classification framework at 10 m resolution for each year. Annual classifications were then combined to create a multi-year frequency layer where pixels with more than two years of continuous cropping were labelled as cropland, creating a stable mask layer. The dataset contains a 10 m resolution binary raster mask layer in GeoTIFF format to support a wide range of applications including cropland mapping, land-use change assessment, agricultural monitoring, yield modelling, soil and climate studies, and machine-learning-based geospatial applications. • Annual Prairie-wide cropland mask cloud-optimized GeoTIFFs generated using AlphaEarth data embeddings and Random Forest models trained on stratified random reference samples. • Multi-year cropland frequency and stable cropland mask layers derived from aggregated annual predictions, enabling consistent identification of persistent cropland across the Canadian Prairies.},
}

@article {pmid42306387,
year = {2026},
author = {De Alba Solis, AU and Gómez Sánchez, E},
title = {CliniCAM: A Technical Report of a Mobile Health Application for Structured Clinical Image Documentation and Tag-Based Dataset Generation.},
journal = {Cureus},
volume = {18},
number = {5},
pages = {e108942},
pmid = {42306387},
issn = {2168-8184},
abstract = {Clinical image recording is often poorly standardized. The use of personal devices and messaging platforms results in data fragmentation and limited accessibility. Previous mobile solutions prioritized secure image capture and electronic health record integration, yet offered minimal support for structured organization and efficient retrieval. This report describes the design and development of CliniCAM, a mobile health application for structured clinical image capture, annotation, tagging, retrieval, and dataset generation. CliniCAM was developed using FlutterFlow (FlutterFlow Inc., Mountain View, CA, USA) with a Firebase backend (Google, Mountain View, CA, USA). Firestore manages structured data, and Firebase Storage handles image management. The application supports in-app image capture, patient association, free-text annotation, and assignment of custom tags. Users can search by patient, free-text, or tag. Data export is enabled through Google Sign-In and the Google Drive API, allowing the generation of datasets containing images and metadata in JSON format. CliniCAM delivers a unified workflow for clinical image documentation by integrating image capture, metadata annotation, and tag-based classification within a mobile interface. The system permits efficient retrieval and supports the creation of tagged datasets for secondary applications. For example, in an orthopedic consultation, a clinician can use CliniCAM to capture high-resolution images of musculoskeletal findings, such as joint deformities, surgical wounds, or traumatic injuries, directly within the app, tag the images with terms such as "osteoarthritis" or "fracture," and enter relevant clinical annotations. These images are immediately associated with the patient record and securely stored. During follow-up visits, the clinician can quickly retrieve previous images using patient identifiers or tags to monitor disease progression. Similarly, in wound care, nurses can document wound healing over time, with images organized by anatomical site and wound type, thereby facilitating clinical decision-making and generating datasets for quality improvement or research. CliniCAM delivers a scalable and affordable solution for structured clinical image documentation. The tag-based system enables dataset generation at the point of care, addressing limitations within traditional storage systems and supporting future research, educational efforts, and artificial intelligence applications.},
}

@article {pmid42307868,
year = {2026},
author = {Abdul Rahim, AI and Moorthy, P and Balu, B and Chokkalingam, L},
title = {Future multi-dimensional drought analysis using machine learning and geospatial approaches: a case study of Tiruchirappalli District, Tamil Nadu, India.},
journal = {Environmental science and pollution research international},
volume = {},
number = {},
pages = {},
pmid = {42307868},
issn = {1614-7499},
abstract = {This research delivers a comprehensive future-oriented multi-dimensional drought appraisal for Tiruchirappalli District, Tamil Nadu, India, by inter-linking Google Earth Engine (GEE) cloud computing, machine learning algorithms, geospatial analysis, and socioeconomic indicators. Four primary drought dimensions-meteorological, hydrological, agricultural, and socioeconomic vulnerability-are interfaced into a holistic drought evaluation framework during 2014-2025, representing an epoch with intensified climate variability in South India. Meteorological drought is assessed through the Standardized Precipitation Index (SPI), while hydrological drought is assessed through the Standardized Water Level Index (SWI) from the observations of the Public Works Department (PWD) groundwater. Agricultural drought conditions are considered using multi-sensor satellite indices on the GEE platform, namely, Vegetation Condition Index (VCI), Temperature Condition Index (TCI), and Precipitation Condition Index (PCI). Socioeconomic vulnerability parameters, which included population density, literacy, household density, and workforce-related characteristics. These socioeconomic indicators were standardized and weighted separately to derive a socioeconomic vulnerability map, while meteorological, hydrological, and agricultural drought indicators were integrated into Multi-Drought Severity Index. The spatio-temporal assessment shows cyclic drought occurrences, which became stronger in the 2015-2018 period and revived from 2021 onward, particularly in the years 2023-2025. LULC analysis with Random Forest classification for 2014, 2018, and 2025 pointed toward rapid urbanization and consequent land-use change, which increases drought vulnerability. Future drought manifestation for 2035 was done by employing multi-year geospatial trends, historical RF-based LULC spatio-temporal change analysis, ANN-based LULC-2035 predictions, and long-term drought indicators. The integrated ANN-based LULC-2035 and MDSI-2035 analysis predicts extreme drought in Thuraiyur, Omandhur, and Thachankurichi, with varying severity across other regions. This study demonstrates the successful application of multi-indicator drought modelling together with machine-learning-driven land-cover prediction, thereby presenting a scalable framework for regional drought risk assessment and climate-resilient planning.},
}

@article {pmid42014779,
year = {2026},
author = {Gupta, A and Gupta, S and Sharma, S and Singh, J and Ali, F and Kwak, D},
title = {A privacy preserving optimized intelligent security framework for smart homes using zero trust architecture and explainability.},
journal = {Scientific reports},
volume = {16},
number = {1},
pages = {},
pmid = {42014779},
issn = {2045-2322},
abstract = {Modern smart home environments include traditional sensors, autonomous robots, connected vehicles, intelligent consumer electronics devices, and wearables, along with context-specific computing and services. The heterogeneity of smart devices, together with their communication protocols, applications, and cloud-hosted services, introduces new vulnerabilities and raises security and privacy concerns. End-to-end foolproof security and data privacy guarantees are currently difficult to achieve and remain an open challenge. This paper proposes POIZE (Privacy, Optimization, Intelligence, ZTA, and Explainability), a comprehensive security framework for modern smart homes. POIZE focuses on preserving privacy, optimizing data flow and processing, using Tiny Machine Learning (TinyML) models for anomaly detection, applying Zero Trust Architecture (ZTA) principles, and providing explainability to end users. The proposed framework combines ZTA principles with novel data processing and computation optimization, while emphasizing explainability, to meet the evolving needs of users and address key security and privacy challenges in smart home environments. Preliminary experimental results demonstrate the feasibility of the proposed framework.},
}

@article {pmid42299169,
year = {2026},
author = {Ajit, A and Reddy, V},
title = {Slow Harm After Dobbs v. Jackson Women's Health Organization: An Analytic Scoping Review of How Data and Surveillance Infrastructures Reshape Emergency Obstetric Care.},
journal = {Cureus},
volume = {18},
number = {5},
pages = {e108911},
pmid = {42299169},
issn = {2168-8184},
abstract = {Following Dobbs v. Jackson Women's Health Organization on June 24, 2022, abortion criminalization in the United States has converged with interoperable health data systems, making reproductive health documentation more portable and legally legible across jurisdictions. This analytic scoping review, guided by Systematic Reviews and Meta-Analyses Extension for Scoping Reviews (PRISMA-ScR), examined U.S. literature published from June 2022 through October 2025 in PubMed, OpenAlex, Google Scholar, and HeinOnline. Sources were organized by evidentiary role into Tier C (legal and analytic conditions), Tier B (organizational adaptations), and Tier A (clinical impacts). The review mapped how surveillance pathways, including electronic health record and health information exchange interoperability, cloud hosting, billing and claims systems, and non-Health Insurance Portability and Accountability Act (HIPAA) consumer and law enforcement data, expand perceived liability and drive documentation and communication chill. Reported responses included vague documentation, reduced data capture, segmented records, and restricted information exchange. These adaptations may fragment care and delay intervention for emergency obstetric conditions. Tier A evidence included delays and higher diagnostic thresholds in ectopic pregnancy management, as well as increased maternal morbidity under mandated expectant management of previable preterm prelabor rupture of membranes and related second-trimester complications. These harms disproportionately affect patients experiencing structural racism, poverty, and intensified surveillance. We argue that these pathways constitute a form of slow harm, in which cumulative injury is produced through delay, fear, and institutional paralysis. Privacy-protective, equity-centered data governance and clearer clinical-legal guidance are needed to preserve safe documentation and coordinated emergency obstetric care.},
}

@article {pmid42056220,
year = {2026},
author = {Sookhtsaraei, R and Sakhaei-Nia, M and Parand, FA},
title = {A modular framework for mission-critical services in fog-based remote healthcare systems: identification, categorization, and evaluation.},
journal = {Scientific reports},
volume = {},
number = {},
pages = {},
doi = {10.1038/s41598-026-50817-3},
pmid = {42056220},
issn = {2045-2322},
abstract = {This paper addresses the challenges of integrating remote healthcare systems that rely on hybrid computing models—such as cloud computing, fog computing, and the Internet of Things (IoT)—to ensure real-time monitoring and reliable data communication. The authors propose a modular, flexible, and dependable system model for mission-critical remote healthcare services, focusing on five key components: Input/output (I/O) services, Artificial intelligence (AI) services, data services, security and privacy services, and context-aware services. Each component supports real-time responsiveness and interoperability, minimizing potential disruptions. A case study on trauma patient care demonstrates the model’s practical application within a hybrid computing framework. The evaluation, based on a simulation setup, shows that the hybrid model reduces latency, underlining the importance of distributed architectures for timely and reliable emergency care. The paper also reviews existing literature, highlighting the model’s versatility and its potential to enhance the delivery of remote healthcare services across various domains, thereby improving overall care quality and system efficiency.},
}

@article {pmid42289574,
year = {2026},
author = {Alfattani, R and Hotami, MM},
title = {Digital twin-driven multiscale modelling for real-time defect prediction in metal additive manufacturing.},
journal = {Scientific reports},
volume = {},
number = {},
pages = {},
doi = {10.1038/s41598-026-58348-7},
pmid = {42289574},
issn = {2045-2322},
abstract = {This paper presents a multiscale modelling system based on a digital twin that can predict defects in metal additive manufacturing in real time, with primary validation scoped to Laser Powder Bed Fusion (LPBF). While the framework is architected to generalise across powder-bed and directed-energy deposition (DED) processes, all experimental evaluations are conducted on LPBF using the publicly available NIST AM-Bench benchmark dataset of IN625 and Ti-6Al-4 V specimens. It combines microscale melt pool dynamics with mesoscale thermal fields and macroscale structural deformation via hierarchical physics-informed neural network (PINN) surrogates, with a real-time Internet of Things (IoT) backbone of sensors. It has a three-tier architecture of edge, fog, and cloud, with inference distributed across latency-sensitive edge nodes, fog-level surrogate aggregation, and cloud-based digital twin calibration. Thermal imaging, acoustic emission, and optical monitoring provide sensor-based feedback to the numerical model, which can be used to adaptively control the process in real time during Laser powder bed fusion (LPBF) and directed energy deposition (DED). The hybrid Convolutional Neural Network-Long Short-Term Memory (CNN-LSTM) is an extractor of spatio-temporal defect signatures used to classify six defect classification categories, including porosity, lack-of-fusion, cracking, balling, keyholing, and delamination, with a macro-averaged F1-score of 0.9841. Formation of the multiscale coupling, Bayesian calibration and surrogate training formalised in twenty-five governing equations. It was experimentally verified on the publicly available NIST AM-Bench dataset that the proposed DT-MSM structure achieves a mean defect-detection rate of 98.72% and an inference latency of 11.3ms per frame on edge hardware, outperforming seven other baselines. The framework achieves reductions in scrap rate (34.6 per cent) and predictive maintenance lead time (41.2 per cent) in a simulated LPBF production scenario, with improved predictive ability for porosity and cracking compared with offline simulations. Mean plus standard deviation of the results is reported across five random seeds, over which the statistical significance is verified using a paired t-test (p < 0.01). The proposed methodology supports smart manufacturing and quality control in the new-generation production systems for metal additive manufacturing.},
}

@article {pmid42291889,
year = {2026},
author = {Bardarov, S and Zarineh, A},
title = {A Reproducible Protocol and Framework for Large Language Model (LLM)-Assisted Estrogen and Progesterone Receptor (ER/PR) Scoring.},
journal = {Cureus},
volume = {18},
number = {5},
pages = {e108708},
pmid = {42291889},
issn = {2168-8184},
abstract = {Estrogen and progesterone receptor (ER/PR) scoring in breast cancer is vulnerable to interobserver variability, particularly at diagnostic thresholds. While large language models (LLMs) offer potential as ancillary tools, no standardized deployment framework exists. This is a feasibility and protocol-development study that establishes a reproducible protocol for LLM implementation in pathology labs. Using College of American Pathologists (CAP) proficiency-testing tissue microarrays, we developed a systematic framework that combines recursive prompt engineering and zero-state reset (ZSR) protocols, requiring fresh chat sessions for each evaluation to prevent conversational bias. Three models were evaluated: Claude Haiku 4.5 (Anthropic, San Francisco, CA, USA), Gemini 3.0 (Google, Mountain View, CA, USA), and Gemma 3.0 12B (Google, Mountain View, CA, USA). The ZSR protocol proved essential, improving accuracy from a 70-80% baseline to ≥95% CAP concordance. Performance varied across models and scoring approaches: Claude achieved the highest concordance (90-98%), followed by Gemini (85-100%) and Gemma (73-93%), demonstrating that systematic prompt refinement, not model capacity, drives diagnostic accuracy. Final concordance exceeded 83% across all models. The locally hosted model approached cloud-level performance within a clinically meaningful range, achieving concordance rates that, while numerically lower, remained within acceptable feasibility thresholds for a protocol development context. We identified systematic error sources, including image artifacts, slide contamination, and compression effects. This protocol provides pathology laboratories with evidence-based guidance for implementing LLM-assisted ER/PR scoring without requiring specialized infrastructure or extensive computational resources. The framework is immediately actionable as a research and quality-assurance tool and provides a reproducible foundation for future clinical validation studies across laboratory settings.},
}

@article {pmid42293526,
year = {2026},
author = {Peng, Y and Jiang, Y and Lee, YJ and Lee, JH and Kim, CY and Lee, J},
title = {TaxaScope: a container-native, visualization-centric workstation for genome-based bacterial taxonomy.},
journal = {Frontiers in microbiology},
volume = {17},
number = {},
pages = {1809734},
pmid = {42293526},
issn = {1664-302X},
abstract = {INTRODUCTION: Genome-based bacterial taxonomy requires standardized and reproducible analytical workflows for species delineation and phylogenomic placement; however, the practical deployment of these workflows remains a significant barrier for experimental biologists and clinical scientists. Widely adopted tools such as Prokka, antiSMASH, and PhyloPhlAn underpin key steps in genome annotation, functional characterization, and phylogenomic reconstruction, but their practical deployment in routine laboratory settings, especially on Windows based systems, remains non trivial due to complex software dependencies and command line centric workflows. Existing solutions, including cloud-based platforms (e.g., Galaxy and KBase) and commercial software suites (e.g., CLC Genomics Workbench), partially alleviate these challenges but may also involve considerations related to data-privacy concerns, upload latency, storage quotas, shared computing resources, and recurring licensing costs.

METHODS: To address these limitations, we introduce TaxaScope, a graphical-interface-driven desktop workstation designed to support reproducible, genome-based bacterial taxonomy by integrating a curated set of community-validated tools for genome quality assessment, annotation, phylogenomic inference, genome relatedness estimation, and functional profiling within a unified local graphical user interface (GUI). By leveraging Docker- and Podman-based containerization behind a user-friendly frontend, TaxaScope provides version-locked, standardized execution environments across computing platforms without requiring manual dependency management or prior Linux expertise.

RESULT AND DISCUSSION: We demonstrate the utility of TaxaScope through a comprehensive re-analysis of Pseudomonas putida KCTC 1751T, illustrating how standardized taxonomic workflows can be executed locally while automatically generating high-quality circular genome maps and interactive functional reports suitable for downstream interpretation and figure preparation directly from native tool outputs. Collectively, TaxaScope lowers the technical barrier to standardized and reproducible genome-based bacterial taxonomy by providing a private, locally controlled, containerized workflow that complements cloud-based and commercial infrastructures for routine taxonomic research. By providing a containerized and visualization-oriented desktop environment, TaxaScope facilitates the standardized execution of established genomic tools, thereby bridging the gap between complex bioinformatic workflows and consistent bacterial taxonomy.},
}

@article {pmid42280939,
year = {2026},
author = {Sun, Y and Shang, Z and Yang, M and Meng, F and Mu, M and Yan, H},
title = {GeoRescue: A Geometric LiDAR Point Cloud Registration Framework for Resource-Constrained Edge Platforms.},
journal = {Sensors (Basel, Switzerland)},
volume = {26},
number = {11},
pages = {},
doi = {10.3390/s26113422},
pmid = {42280939},
issn = {1424-8220},
support = {202602030088SF//Jilin Province Science and Technology Department/ ; },
abstract = {Accurate LiDAR point cloud registration on resource-constrained edge platforms is a prerequisite for intelligent robotics and industrial automation, yet it remains challenging because low-overlap matching, false correspondences, and fine alignment must be handled under limited computing budgets without GPU acceleration. While learning-based methods have advanced the field, their heavy hardware dependency and training requirements often hinder their practical deployment on mobile edge devices. To bridge this gap, this paper proposes GeoRescue, a training-free geometric registration framework designed for high-precision perception under stringent hardware limits. The method consists of three modular stages: Asymmetric Correspondence Expansion (ACE), which enlarges the candidate correspondence set to reduce the loss of true matches; Dynamic Geometric Topology Gating (DGTG), which suppresses false matches through distance-consistency-based hypothesis filtering; and Uncertainty-Aware Manifold Refinement (UAMR), which improves fine alignment by explicitly modeling local anisotropic noise via covariance-guided optimization. Experiments on 3DMatch, 3DLoMatch, and KITTI show that GeoRescue achieves registration recall rates of 84.84% and 41.27%, respectively, and a 94.95% success rate on KITTI. Remarkably, the framework matches the accuracy of high-capacity learning models while running on a GPU-free, 15 W edge CPU platform (Intel Core i5-8265U). These results indicate that GeoRescue provides a deployment-ready solution with an optimal efficiency-accuracy trade-off for LiDAR sensing and robotics perception in complex, real-world scenarios.},
}

@article {pmid42281033,
year = {2026},
author = {Wiangnak, V and Wiboonrat, M and Duangsuwan, S},
title = {Development of an AIoT-Based Early Flash-Flood Warning System for Smart Rural Disaster Resilience.},
journal = {Sensors (Basel, Switzerland)},
volume = {26},
number = {11},
pages = {},
doi = {10.3390/s26113512},
pmid = {42281033},
issn = {1424-8220},
support = {Grant number 2566-02-08-004//King Mongkut's Institute of Technology Ladkrabang/ ; },
abstract = {This paper presents the development of an AIoT-based early flash-flood warning system to enhance disaster resilience in smart rural communities. The framework integrates multi-source hydrological sensors, AI-enabled edge-cloud computing, and a mobile alert application to provide real-time monitoring and short-term flood forecasting, and includes an intelligent hybrid model combines YOLOv10 for visual water-level detection from CCTV imagery with a long short-term memory (LSTM) network for hydrological time-series prediction. The system was deployed and evaluated at two sites in Thailand: the Ban Luang station in Chiang Mai and the Chumkho station in Chumphon. The experimental results show near-perfect detection performance by YOLOv10, with precision and mAP@0.5 exceeding 0.99 across varying water-level conditions. The LSTM model achieved high forecasting accuracy, with an R[2] of 0.987 at Ban Luang and 0.781 at Chumkho, reflecting site-specific hydrodynamic complexity. The results confirm that integrating AIoT-based visual sensing with data-driven forecasting significantly improves the reliability, responsiveness, and robustness of early flash-flood warning systems in rural environments.},
}

@article {pmid42286048,
year = {2026},
author = {Dixit, RS and Choudhary, SL and Arya, N and Nathani, N and Bhowmik, C and Roy, V and Jain, S},
title = {Artificial intelligence-powered cloud security strategies for protecting critical clinical operations in healthcare environments.},
journal = {Scientific reports},
volume = {},
number = {},
pages = {},
doi = {10.1038/s41598-026-55522-9},
pmid = {42286048},
issn = {2045-2322},
abstract = {The increasing use of cloud computing in hospitals, telemedicine, the Internet of Medical Things (IoMT) and real-time patient monitoring has made for an increasing trend of artificial intelligence-driven cloud security in hospitals. The growing reliance on distributed healthcare clouds layers the cyber-attack surface, however, with critical clinical operations now at risk from ransomware attacks, insider threats, API exploitation, and advanced persistent attacks. This research study introduces a novel AI-integrated cloud security framework tailored for safeguarding mission critical applications in the healthcare sector featuring an intelligent threat detection component, a probabilistic risk evaluation system, and an adaptive response orchestration system. The proposed architecture is built-in by using telemetry normalization, probabilistic behaviour modelling, deep autoencoders for anomaly detection, Bayesian approach for threat probability estimation, multi-objective risk scoring and reinforcement learning for adaptive mitigation. An experimental validation was performed with the CICIDS2017 dataset including around two million samples of network traffic data across various attack categories. The experimental results show that excellent performances have been achieved with an accuracy of 0.96, precision of 0.95, recall of 0.94, F1score of 0.95 and AUC of 0.98 with low latency of around 26 ms and reduced false positive rate of 0.03. A comparative analysis against the current cloud security methods also confirms the effectiveness of the proposed framework in delivering better operational security, response time and the availability of clinical services, providing the continuous clinical service that healthcare organizations require. The research showcases how incorporating AI with responsive cloud security mechanisms can offer a scalable and resilient defense against today's healthcare cloud infrastructures.},
}

@article {pmid42270792,
year = {2026},
author = {Alkhattabi, K and Belhaj, S and Selecky, J and Talha, M},
title = {An adversarial-resilient intrusion detection framework for internet of medical things (IoMT) using digital twin-enabled behavioral threat modeling and federated hybrid ensemble learning.},
journal = {Scientific reports},
volume = {},
number = {},
pages = {},
doi = {10.1038/s41598-026-55893-z},
pmid = {42270792},
issn = {2045-2322},
support = {NBU-FFR-2026-2200-01//Northern Border University/ ; },
abstract = {The Internet of Medical Things (IoMT) has transformed healthcare by enabling continuous patient monitoring and remote diagnostics. However, this growth introduces considerable security challenges. This paper examines vulnerabilities in IoMT devices to advanced cyberattacks that jeopardize patient safety and data integrity. We review the limitations of traditional security methods and motivate the need for adaptive defenses. We evaluate machine learning and deep learning models for real-time threat detection and identification of anomalous behavior within IoMT networks. We further propose a security framework that integrates digital twin technology with edge-cloud computing to improve the reliability of IoMT applications. Results show that hybrid and deep-learning models maintain detection performance under the resource constraints typical of medical devices. The proposed XGBoost component achieved a precision of 0.97, a recall of 0.98, and an ROC-AUC of 0.999 on the SmartWard dataset, while the hybrid ensemble showed measurable adversarial robustness under FGSM perturbations. The decision-critical inference path runs in under 0.05 seconds, supporting deployment on resource-constrained medical devices.},
}

@article {pmid42272483,
year = {2026},
author = {Zhao, B and Li, J and Qiao, C and Liu, JN and Pei, Q and Shen, Y},
title = {TRUST: A toolkit for TEE-assisted secure outsourced computation over integers.},
journal = {Fundamental research},
volume = {6},
number = {3},
pages = {1816-1826},
pmid = {42272483},
issn = {2667-3258},
abstract = {Secure outsourced computation (SOC) provides secure computing services by taking advantage of the computation power of cloud computing and the technology of privacy computing (e.g., homomorphic encryption). Expanding computational operations on encrypted data (e.g., enabling complex calculations directly over ciphertexts) and broadening the applicability of SOC across diverse use cases remain critical yet challenging research topics in the field. Nevertheless, previous SOC solutions frequently lack the computational efficiency and adaptability required to fully meet evolving demands. To this end, in this paper, we propose a toolkit for TEE-assisted (Trusted Execution Environment) SOC over integers, named TRUST. In terms of system architecture, TRUST falls in a single TEE-equipped cloud server only through seamlessly integrating the computation of REE (Rich Execution Environment) and TEE. In consideration of TEE being difficult to permanently store data and being vulnerable to attacks, we introduce a (2, 2)-threshold homomorphic cryptosystem to fit the hybrid computation between REE and TEE. Additionally, we carefully design a suite of SOC protocols supporting unary, binary and ternary operations. To achieve applications, we present SEAT, secure data trading based on TRUST. Security analysis demonstrates that TRUST enables SOC, avoids collusion attacks among multiple cloud servers, and mitigates potential secret leakage risks within TEE (e.g., from side-channel attacks). Experimental evaluations indicate that TRUST outperforms the state-of-the-art and requires no alignment of data as well as any network communications. Furthermore, SEAT is as effective as the Baseline without any data protection.},
}

@article {pmid42265297,
year = {2026},
author = {Azarnoush, S and Nemati, A},
title = {Transportation 4.0 planning in emergency medical services considering real-time ambulance two-phase assignment-routing and mission change.},
journal = {Scientific reports},
volume = {},
number = {},
pages = {},
doi = {10.1038/s41598-026-56495-5},
pmid = {42265297},
issn = {2045-2322},
abstract = {Transportation 4.0 enables smart planning in urban and interurban transportation networks by leveraging intelligent technologies, such as Internet of Things (IoT), cloud computing, information system integration, and artificial intelligence, to facilitate real-time vehicle assignment and routing. Nevertheless, Transportation 4.0 planning involving vehicle smart assignment and routing, particularly in healthcare logistics, has been less addressed in the literature. However, this study takes an initial step toward intelligent planning of emergency medical services by proposing an Intelligent Emergency Information System (IEIS) for real-time ambulance assignment and routing. As part of data analytics in the designed IEIS, an integer linear mathematical model is proposed to optimize total travel times in the ambulance assignment and routing problem based on deterministic real-time input data in two phases: reaching the patients and transporting them to hospitals. The proposed mathematical model is solved in a case study of Amol city, Iran, incorporating 25 street nodes, nine ambulance nodes, six call nodes, and five hospital nodes, using the CPLEX solver. Although designing and implementing the suggested conceptual model of IEIS was not applicable in the case study, to evaluate the impact of real-time parameter updates, two scenarios involving real-time inclusion of a new call node and an ambulance breakdown are examined. Accordingly, the model was re-solved, producing updated ambulance assignments and routing in each scenario, even by rearranging the ambulances' missions. Results validated the flexibility of the proposed mathematical model in optimizing two-phase ambulance assignment and routing under real-time input updates.},
}

@article {pmid42260691,
year = {2026},
author = {Ferro Desideri, L and Bernardi, E and Troyas, C and Subhi, Y and Sabaner, MC and Zinkernagel, M and Anguita, R},
title = {Code-free automated machine learning for OCT-based classification of vitreoretinal interface diseases.},
journal = {International journal of retina and vitreous},
volume = {},
number = {},
pages = {},
doi = {10.1186/s40942-026-00880-9},
pmid = {42260691},
issn = {2056-9920},
abstract = {BACKGROUND: Differentiation of vitreoretinal interface disorders on optical coherence tomography (OCT) relies on expert interpretation and can be challenging in borderline cases. Automated machine learning (AutoML) platforms may enable clinician-driven artificial intelligence development without coding expertise. This study evaluated the performance of a code-free AutoML approach for OCT-based classification.

METHODS: In this cross-sectional image classification study, 434 OCT B-scans from publicly available datasets were manually labeled into four categories: epiretinal membrane (ERM), lamellar macular hole (LMH), full-thickness macular hole (MH), and normal retina. Images were uploaded to a cloud-based AutoML platform (Google Cloud Vertex AI), which automatically performed data splitting (80% training, 10% validation, 10% test), model training, and optimization. Performance was assessed using precision, recall, average precision, and confusion matrix analysis.

RESULTS: The model achieved an overall average precision of 0.988, with precision and recall of 97.6%. MH and normal retina were classified with perfect precision and recall (100%). ERM showed high precision (100%) with slightly reduced recall (92.9%), while LMH demonstrated complete recall (100%) with lower precision (83.3%). Misclassifications were limited to anatomically related entities.

CONCLUSIONS: Code-free AutoML enables accurate OCT-based classification of vitreoretinal interface disorders using a clinician-driven workflow. This approach may facilitate broader adoption of artificial intelligence in ophthalmology and support rapid clinical research prototyping.},
}

@article {pmid42261408,
year = {2026},
author = {Fendzi Mbasso, W and Harrison, A and Mamadiyarov, Z and Liu, Z and Kumar, R and Shaikh, MS and Metwally Mahmoud, M},
title = {Scalable HMO-CNN-SVM Framework for Skin Lesion Classification: A Metaheuristic-Driven Approach With Parallelizable Optimization for Cluster Deployment.},
journal = {Biomedical engineering and computational biology},
volume = {17},
number = {},
pages = {11795972261453621},
pmid = {42261408},
issn = {1179-5972},
abstract = {In medical image analysis, accurate skin lesion categorization is still a major difficulty particularly under limited data conditions and computational complexity. For automated skin cancer detection, in this work we present a scalable hybrid model combining a Convolutional Neural Network (CNN), the Harmonic Mean Optimizer (HMO), and a Support Vector Machine (SVM) classifier-termed HMO-CNN-SVM. Key CNN hyperparameters including learning rate, batch size, and kernel configuration are optimized using the HMO, so greatly boosting classification performance over manual or stationary settings. The model further uses SVM on CNN feature embeddings modified on HMO to improve decision boundary sharpness. Robust performance is shown by experiments carried out on the ACS skin lesion dataset validated by 5-fold cross-valuation and ISIC 2018 benchmarks with an accuracy of 95.02% and consistent generalizing over folds. Crucially, significant parallelism potential made possible by the population-based structure of HMO makes the framework fit for GPU clusters or cloud-based training pipelines. Computational benchmarks expose reasonable overhead in trade for best performance. Thus, the suggested system is a strong contender for implementation in high-performance and distributed computing contexts since it provides both diagnostic dependability and computational tractability.},
}

@article {pmid42252024,
year = {2026},
author = {Rebmann, AW and Okpanum, J and Ghosh, S and Mergy, M and Przybycien, TM and Bequette, BW},
title = {Smart manufacturing for continuous downstream processing of monoclonal antibodies at the lab-scale.},
journal = {SLAS technology},
volume = {},
number = {},
pages = {100441},
doi = {10.1016/j.slast.2026.100441},
pmid = {42252024},
issn = {2472-6311},
abstract = {Continuous bioprocessing has shown promise for efficiently scaling production of monoclonal antibodies for biotherapeutic production. Integrating process analytical technologies at the lab-scale stage improves the reliability of results during development and the translation of the system during scale-up. Smart Manufacturing provides a framework for implementing process analytical technologies at all process scales by leveraging modern hardware, software, and techniques. Demonstration of scaling and retrofitting a legacy system into a smart manufacturing system for the continuous downstream processing of monoclonal antibodies at the lab-scale was achieved through modular and flexible controller hardware, Industrial Internet-of-Things architecture, interoperable OPC-UA servers, and cloud computing on the Smart Manufacturing Interoperability Platform. The system consisted of lab-scale equipment connected to a legacy control device with limited operation and I/O. A new system of hardware was implemented that interfaced with the legacy equipment while also providing scalability, reconfigurability, and controllability. Peristaltic pumps, weigh-scales, single-use pressure sensors, and optical sensors were connected to edge devices which bridged the hardware and software used for process monitoring and control. Lab-scale implementation and operation challenges such as connecting multiple vendor systems and inconsistent pump flowrates were addressed. Water-based pump flowrate control tests were performed on the legacy equipment with readiness for deployment onto the smart manufacturing system. The techniques used include a Kalman Filter, steady-state data reconciliation, and closed-loop control based on scale measurements.},
}

@article {pmid42245193,
year = {2026},
author = {Parra-Royón, M and Garrido-Sánchez, J and Sánchez-Expósito, S and Darriba-Pol, L and Sánchez-Castañeda, J and Mendoza, MÁ and Coles, J and McConkey, S and Joshi, R and Barnsley, R and Salgado, J and Verdes-Montenegro, L},
title = {Towards a sustainable astronomical data infrastructure: Optimising linking data from the Rucio datalake to the users areas within the SKA Regional Centres Network.},
journal = {Open research Europe},
volume = {6},
number = {},
pages = {18},
pmid = {42245193},
issn = {2732-5121},
abstract = {The distributed architecture of the SKA Regional Centre Network (SRCNet) aims to provide scientific communities worldwide with efficient computational and storage resources to exploit the massive data volumes produced by the SKA Observatory (SKAO). Given the amount of SKAO data, traditional data management paradigms - where data is transferred to computational resources- are no longer feasible. Instead, computational workflows must increasingly be relocated closer to data storage locations, emphasizing efficient data access strategies and avoiding unnecessary duplication or redundancy. In this context, we present PrepareData, a modular and extensible data delivery service developed within SRCNet prototyping activities. Our proposal for this service addresses the critical challenge of redundant data transfers and duplication at both node and user levels by enabling seamless delivery of requested datasets from local Rucio Storage Elements (RSEs) directly into users' working environments. PrepareData operates as a local service within each SRCNet node and it is integrated into a broader ecosystem of federated services. Specifically, we designed and evaluated two distinct yet complementary implementations to avoid unnecessary data duplication and to enable a dynamic data bridge between the RSEs and the user storage areas, through: (1) a filesystem-based solution leveraging CephFS, which uses shared filesystem mount points and bind mounts to ensure consistent and immediate data availability of the data across computational nodes, and (2) a Kubernetes model using Persistent Volumes and Persistent Volume Claims, dynamically injecting data into a user's areas. To tackle this work we detail the architectural design and development, the technical implementation, the integration of both solutions with science enabling tools, such as JupyterHub, CARTA or virtually any application, and finally we provide a performance evaluation. This contribution provides a scalable and sustainable blueprint for data delivery in federated scientific infrastructures, supporting the broader goals of green computing and efficient resource utilisation.},
}

@article {pmid42245277,
year = {2026},
author = {Zhang, Y and Reyes-Muñoz, P and Verrelst, J},
title = {A cloud-computing framework for downscaled global 300 m SIF retrieval from Sentinel-3 and TROPOSIF.},
journal = {International journal of applied earth observation and geoinformation : ITC journal},
volume = {150},
number = {},
pages = {105330},
pmid = {42245277},
issn = {1569-8432},
abstract = {Sun-induced chlorophyll fluorescence (SIF) is a critical indicator of photosynthetic activity. Yet, existing satellite SIF products typically suffer from coarse spatial resolutions, generally coarser than 500 m, which limits their utility for fine-scale ecosystem studies. Here, we present a cloud-computing framework designed for the generation of a downscaled SIF product (S3-SIF743) derived from Sentinel-3 (S3) Ocean and Land Colour Instrument (OLCI), with a spatiotemporal resolution of 300 m and 4 days. Our approach uses the Google Earth Engine (GEE) cloud-computing platform to integrate SIF produced from TROPOspheric Monitoring Instrument measurements within the 743-758 nm retrieval (TROPOSIF 743), S3 radiances, S3-based vegetation traits, latitude and longitude within a Random Forest (RF) regression framework. Model training over Europe achieved robust performance against TROPOSIF 743 reference data (R 2 = 0.767, RMSE = 0.137 mW m[-2] sr - 1 nm - 1), and the approach was subsequently extended globally. Validation against ground-based tower observations confirmed that S3-SIF743 effectively reproduces seasonal dynamics across diverse ecosystems. Comparisons against TROPOSIF 743 demonstrated strong spatial consistency in temperate agricultural regions, with the highest R 2 values in croplands and lower agreement in sparsely vegetated or persistently cloudy regions. Global mapping revealed coherent patterns of photosynthetic activity, with peak values in tropical rainforests and major agricultural zones. Importantly, S3-SIF743 reduces retrieval noise relative to TROPOSIF 743 and provides unprecedented insights into sub-kilometer spatial heterogeneity. By combining S3's rich spectral capabilities with GEE's scalable computing environment, our approach bridges the gap between current coarse-resolution SIF products and ESA's upcoming FLEX mission, offering a flexible and operational pathway for high-resolution monitoring of terrestrial photosynthesis.},
}

@article {pmid42239803,
year = {2026},
author = {Dozzo, M and Aiuppa, A and Bilotta, G and Cappello, A and Ganci, G},
title = {Volcanic SO2 total mass dataset on Mt. Etna (Italy) from 2018 to 2025 using sentinel-5P TROPOMI.},
journal = {Data in brief},
volume = {66},
number = {},
pages = {112876},
pmid = {42239803},
issn = {2352-3409},
abstract = {Sulfur dioxide (SO2) is released by the magma degassing in the shallow crust, constituting a key indicator of magma ascent rates in the feeding conduit, as well as providing information on the style and the intensity of eruptive activity. Continuous monitoring of this gas is important to understand volcanic processes and to contribute to hazard assessment. The dataset presented here provides a comprehensive time series of SO2 total mass from Mount Etna (Sicily, Italy), covering the period from 2018 to 2025. The data have been obtained from the TROPOspheric Monitoring Instrument (TROPOMI) onboard the Sentinel-5 Precursor satellite, which has been operational since 2018, delivering atmospheric column measurements of sulfur dioxide and other gases at unprecedented spatial resolution and daily revisit time. Volcanic SO2 plumes were automatically identified through a two-step procedure: firstly, the Simple Non-Iterative Clustering (SNIC) segmentation method was applied, which is an object-based image analysis technique and secondly, K-means unsupervised clustering was used on the segmented imagery to further improve cloud detection. The algorithm has been implemented in the open-source Google Earth Engine platform, enabling efficient processing of the TROPOMI imagery collection, to which quality control filters are already applied. This methodological framework supports the generation of SO2 total mass time series with reduced delay and improved calculation time, thus providing a valuable tool for rapid and reliable monitoring of volcanic emissions and for enhancing volcanic hazard assessment capabilities.},
}

@article {pmid42243174,
year = {2026},
author = {Tripathy, N and Sahoo, S and Alghamdi, NS and Dhiman, G},
title = {Hybrid GA-DQL approach for efficient task mapping of IoT applications in fog computing framework.},
journal = {Scientific reports},
volume = {},
number = {},
pages = {},
doi = {10.1038/s41598-026-50629-5},
pmid = {42243174},
issn = {2045-2322},
abstract = {Fog computing has emerged as a promising paradigm to extend cloud services closer to IoT devices, improving response times & reducing network congestion. However, efficient load balancing in fog computing is essential to maximize performance, reduce costs, ensure energy efficiency, & maintain a high quality of service, ultimately supporting the demands of latency-sensitive & resource-intensive IoT applications. The primary objective of task mapping in computing environments such as cloud, fog, & edge computing is to allocate tasks across available resources in an efficient & effective manner, particularly in fog computing, where resources are distributed & closer to end devices. This paper presents a hybrid approach that integrates Genetic Algorithm (GA) & Deep Q-Learning (DQL) for task mapping in fog computing environments. The objective is to minimize makespan & computational costs while maintaining high resource utilization. Our approach leverages a GA to perform initial task allocation by exploring a broad solution space, thereby enhancing convergence toward optimal scheduling patterns. This solution is refined using DQL, which adapts to dynamic environments by learning from continuous feedback and enabling real-time decision-making. By combining the exploration strengths of GA with the adaptive capabilities of DQL, the proposed method effectively manages task allocation & resource utilization. Experimental results show that our hybrid approach outperforms baseline methods, significantly reducing makespan & operational costs.},
}

@article {pmid42243474,
year = {2026},
author = {Alwabli, A},
title = {Optimized multi-tier task offloading strategy for sustainable IoV systems in 6G networks.},
journal = {Scientific reports},
volume = {},
number = {},
pages = {},
doi = {10.1038/s41598-026-56588-1},
pmid = {42243474},
issn = {2045-2322},
abstract = {Sixth-generation (6G) networks are likely to support advanced Internet of Vehicles (IoV) applications that have rigid latency, reliability, and computation demands. Nevertheless, efficient task offloading is a challenging problem because vehicle environments are characterized by mobility, changing channels, varying task requirements, constrained edge resources, and growing energy demands. To address these issues, this study presents an Optimized Multi-Tier Task Offloading Strategy (OMTOS) for sustainable IoV systems. The proposed framework comprises a four-tier computing architecture comprising vehicles, roadside units (RSUs), mobile edge computing (MEC) servers, and cloud infrastructure. The generalized latency-energy optimization problem is formulated to allocate tasks across these levels, accounting for task due dates, resource capacity, communication delay, computation delay, and energy consumption. To address dynamic offloading, OMTOS employs a centralized training and decentralized execution (CTDE) based multi-agent Soft Actor-Critic (SAC) method, where the vehicle agents can make decentralized offloading decisions with centralized critics guiding the coordinated learning process during training. It is tested against rule-based and heuristic as well as deep reinforcement learning and various multi-agent reinforcement learning baselines, including LE, EO, RO, GO, DQN, DDPG, SAC, MADDPG, and MAPPO. The aforementioned results reveal that OMTOS achieves low average delay, low energy consumption, a high task success rate, and high convergence compared to the competing methods. Sensitivity analysis also indicates that the latency and energy weightings can be changed to suit various IoV service requirements, including delay-critical safety services, and energy-conscious delay-tolerant services. These results show that OMTOS offers an adaptive and sustainable task-offloading tool in 6G-enabled IoV environments.},
}

@article {pmid42231952,
year = {2026},
author = {Soares, IV and Furberg, A and Azizi, S and Finnveden, G and Yarime, M and Klemun, MM},
title = {Estimating cross-border cloud computing emissions: A consumption-based approach applied to major European data center hubs.},
journal = {iScience},
volume = {29},
number = {6},
pages = {116061},
pmid = {42231952},
issn = {2589-0042},
abstract = {Cloud computing's expansion creates a significant carbon footprint from data center energy consumption. Standard location-based accounting attributes cloud emissions to countries where data centers are located, potentially obscuring where actual demand lies when services are consumed across borders. This study develops a consumption-based accounting approach to estimate cross-border cloud emission flows among the EU's three largest data center hubs: Germany, Ireland, and the Netherlands, from 2017 to 2022. Using electricity consumption and emissions for data centers and telecommunications networks, we classify cloud activity as domestic (hosted and consumed within one country), export (domestic hosting, foreign consumption), or import (foreign hosting, domestic consumption). The analysis reveals distinct profiles: Germany's cloud footprint is predominantly domestic (75%), Ireland's is export-dominated (85% serves foreign consumers), and the Netherlands exhibits balanced flows. Consumption-based metrics can complement territorial accounting by revealing cross-border emission flows relevant to national climate planning and infrastructure policy decisions.},
}

@article {pmid41957468,
year = {2026},
author = {Xia, X and Sun, S and Liu, Q and Guo, H and Wang, Y},
title = {Spatio-temporal heterogeneity and driving mechanisms of RSEI in the north-south sections of the Beijing-Hangzhou grand canal: an empirical study using GEE and XGBoost-SHAP.},
journal = {Scientific reports},
volume = {16},
number = {1},
pages = {},
pmid = {41957468},
issn = {2045-2322},
abstract = {The Beijing-Hangzhou Grand Canal, a vital ecological corridor and cultural heritage site, requires a comprehensive understanding of the spatio-temporal evolution and driving mechanisms of its ecological environment to support sustainable regional development. This study leveraged the Google Earth Engine cloud platform and MODIS growing-season imagery (May-September, 2000–2020) to assess the spatiotemporal dynamics of ecological quality along the entire canal using the Remote Sensing Ecological Index (RSEI). An explainable machine learning framework (XGBoost-SHAP) was further applied to quantitatively disentangle the contributions of natural and anthropogenic drivers underlying the observed spatial heterogeneity in RSEI. The results revealed that: (1) A pronounced and persistent north-south gradient in RSEI values was identified, with ecological quality consistently higher in southern regions compared to northern regions over the two-decade period; and (2) the driving mechanisms demonstrated distinct differences between sections-the ecological quality in the northern section was primarily shaped by natural factors such as precipitation and temperature (“natural factor-dominated” regime), whereas in the southern section it was mainly driven by nighttime light intensity, indicative of urbanization (human activity-dominated” regime). This study elucidates the differential causes of ecological quality divergence between the north and south sections of the canal. The integrated GEE and XGBoost-SHAP framework provides a robust and interpretable approach for attribution analysis in complex environmental systems. This approach has the potential to be extended to other large linear ecosystems and provides a scientific basis for region-specific ecological protection and restoration strategies.},
}

@article {pmid42230710,
year = {2026},
author = {Salem, AI and Elbaz, M and Khalil, HM and Loey, M},
title = {SQUID-COMM: a Colossal Squid-inspired distributed communication framework for real-time multi-node aquaculture monitoring networks with adaptive bioluminescent signaling and neuromorphic edge intelligence.},
journal = {Scientific reports},
volume = {16},
number = {1},
pages = {},
pmid = {42230710},
issn = {2045-2322},
abstract = {Precision aquaculture demands robust communication networks capable of coordinating thousands of distributed sensors across marine and freshwater facilities. Current aquaculture IoT networks face critical challenges including underwater signal attenuation reaching 98% loss at 100 m depth, dynamic topology changes from fish movement and water currents, and severe energy constraints on battery-powered sensor nodes. This paper introduces SQUID-COMM, a novel bio-inspired communication framework emulating the signaling mechanisms of the Colossal Squid (Mesonychoteuthis hamiltoni). The framework introduces seven innovative mechanisms: Bioluminescent Pulse-Coded Modulation (BPCM) achieving 34% higher spectral efficiency through adaptive signal encoding; Chromatophore-Inspired Channel Adaptation (CICA) enabling 15ms frequency hopping response time; Distributed Axon-Ganglia Routing Protocol (DAGRP) maintaining 99.7% packet delivery under 40% node mobility; Tentacle-Topology Self-Organization (TTSO) for dynamic mesh network formation; Giant Fiber Emergency Broadcast (GFEB) achieving sub-50ms critical alert propagation; Photophore Synchronization Protocol (PSP) for microsecond-accurate time coordination; and Ink-Cloud Congestion Control (ICCC) reducing packet loss by 82%. The Enhanced SQUID-COMM variant incorporates Neuromorphic Edge Processing reducing cloud communication by 78%, Federated Learning Coordination for distributed model updates, and Quantum-Resistant Encryption for future-proof security. Experimental evaluation across five aquaculture deployment scenarios demonstrates end-to-end latency of 12.3ms representing 78% reduction compared to LoRaWAN, throughput of 2.4 Mbps in turbid conditions spanning 5-150 NTU, energy efficiency of 0.23 mJ/bit constituting 67% improvement over Zigbee, and network lifetime extension of 340%. Real-world deployment at four commercial facilities across Norway, Egypt, Thailand, and Greece over 120 days processed 2.3 billion sensor readings with 99.94% reliability, enabling fish behavior detection at 94.7% accuracy and early disease detection with 4.2-day lead time. Statistical analysis confirms significant improvements with p-values below 0.001 and Cohen's d exceeding 1.2, while economic evaluation demonstrates annual savings of €89,000-€340,000 per facility.},
}

@article {pmid42230914,
year = {2026},
author = {Zhang, X},
title = {Design of an AI-based security anomaly detection system for IoT terminals based on the ViT-transformer fusion model.},
journal = {Scientific reports},
volume = {},
number = {},
pages = {},
doi = {10.1038/s41598-026-55476-y},
pmid = {42230914},
issn = {2045-2322},
abstract = {The deep penetration of IoT terminals in water systems, healthcare, transportation, and other fields has exacerbated security threats such as cyber-physical attacks and traffic anomalies. However, traditional anomaly detection methods have limitations such as dependence on labeled data, weak generalization ability, high resource consumption, and prominent privacy risks. Although Vision Transformer (ViT) has the advantage of capturing global features, it is difficult to directly adapt to resource-constrained IoT terminals. In existing research, hybrid deep learning models have improved detection accuracy, but lightweight ViT fusion models lack terminal adaptability and multi-modal data fusion applications are scarce. The balance between dynamic scheduling and privacy protection in end-edge-cloud collaboration still needs to be broken through. To address the above issues, this paper proposes an IoT terminal AI security anomaly detection system based on the ViT-Transformer fusion model: adopting a three-level end-edge-cloud collaborative architecture, integrating multi-modal data such as network traffic, sensor timing, and side channel signals, and achieving cross-modal feature fusion through tokenization; combining pruning, distillation, and quantization optimization strategies to increase the model compression ratio to 70%; introducing Elliptic Curve Certificateless Encryption (CL-PKE) and Batch Listing Signature (BLS) batch authentication to ensure data security, and using federated learning to aggregate edge model updates and optimize global performance. Experiments were conducted on public datasets such as IoT-23 and UCI, as well as a self-made testbed. The results show that the model achieves an accuracy of 89.2% and an F1-score of 0.87 in multi-modal anomaly detection, with a terminal inference delay of 90ms and a memory footprint of 30MB, adapting to low-computing devices such as RPi4B and Arduino; CL-PKE resists brute force attacks for 5.2e6 seconds, and batch authentication for 100 terminals takes only 75ms; it exhibits excellent generalization across smart home, industrial IoT, and other scenarios, with a defense success rate of 85.3% against FGSM attacks. This study effectively addresses the resource bottleneck and security pain points of existing methods, providing an efficient and reliable technical solution for IoT terminal security.},
}

@article {pmid42222812,
year = {2026},
author = {Pongpom, M and Wattanasombat, S and Aphiwongcharoen, C and Limsamutpet, K and Wangsanut, T},
title = {Bridging Manual and Computational Approaches: The Pseudophosphatase Scanner for Genome-Wide Fungal Pseudophosphatome Analysis.},
journal = {ACS omega},
volume = {11},
number = {20},
pages = {29874-29891},
pmid = {42222812},
issn = {2470-1343},
abstract = {Pseudophosphatases are proteins with mutations in their catalytic motifs, resulting in the predicted loss of enzymatic activity. Although pseudophosphatases are established regulators of various signaling pathways in humans and other metazoans, their biological roles in fungi remain largely unexplored. Identifying fungal-specific pseudophosphatases is particularly important because they control fungal growth, development, and virulence through noncatalytic mechanisms and could represent selective antifungal targets due to the absence of close human homologues. Here, we present a comprehensive overview of fungal pseudophosphatases identified across all major phosphatase families. To compile a robust data set, we integrated information from literature and databases as well as systematically scanned the phosphatomes of the human pathogenic yeast Cryptococcus neoformans, the plant pathogen Fusarium graminearum, and the neglected pathogenic fungus Talaromyces marneffei. We identified candidate pseudophosphatases and assessed their conservation through BLAST analysis across humans, true yeasts, filamentous ascomycetes, and basidiomycetes. We classified pseudophosphatase candidates into 10 distinct groups, including three groups (CC1-type Oca, CC1-type Yvh1, and HAD-type NIF) that appear to be fungal-specific with no human homologues and exhibit lineage-specific features. Available evidence indicates that fungal pseudophosphatases contribute to signaling pathways that regulate development, metabolism, stress responses, and virulence. In addition, we developed the Pseudophosphatase Scanner tool as a post-HMM (Hidden Markov Model) analysis pipeline to enable genome-wide pseudophosphatase detection. By combining HMM scan-based fold assignment with motif-level pattern matching, the Pseudophosphatase Scanner distinguishes canonical motifs, relaxed variants, and degenerated fold remnants, facilitating functional interpretation of phosphatase active sites at the sequence level. The Pseudophosphatase Scanner is accessible both as a Python script for local execution and as a Google Colab notebook, offering a point-and-click interface for cloud-based analysis. This study provides a catalog of fungal pseudophosphatases and a bioinformatics platform for efficient large-scale pseudophosphatase discovery.},
}

@article {pmid42209538,
year = {2026},
author = {Cheng, YT and Hoang, NT and Shinoda, Y and Islam, K and Motoshita, M and Kadoya, T and Kanemoto, K},
title = {Mapping global resource driven nature loss in the mining sector from 2001 to 2022.},
journal = {Nature communications},
volume = {},
number = {},
pages = {},
doi = {10.1038/s41467-026-73792-9},
pmid = {42209538},
issn = {2041-1723},
abstract = {Global mineral extraction is expected to surge due to the growing demand for clean energy. While mining is critical to modern society, its environmental impacts, though increasingly studied, remain undocumented for half of the world's mining areas and are rarely analysed at the commodity level. Here, we introduce a novel approach integrating remote sensing, machine learning, and cloud computing to classify approximately 70,000 mining sites by commodity. Using this newly detailed dataset, we quantify the nature loss associated with 20 extracted commodities, focusing on deforestation and habitat destruction. From 2001 to 2022, mining activities worldwide resulted in the removal of 16,268 km[2] of forest cover, with 65.64% occurring in tropical and subtropical regions. Notably, approximately half of this deforestation was attributed to the extraction of gold, coal, aluminium (bauxite), nickel-cobalt and copper, primarily in countries intersecting the Amazon, Southeast Asian, and Congo Basin rainforests. Our analysis also reveals that deforestation-to-mining area ratios and biodiversity risks vary by mining location, and conservation threats do not always scale with deforestation rates. By providing commodity-specific maps of mining-induced nature loss, our work equips companies and organisations with actionable insights to identify risks within their supply chains and implement targeted mitigation strategies.},
}

@article {pmid42211407,
year = {2026},
author = {Churches, TR and Green, R and Nguyen, PB and Shawon, MSR and Jorm, LR},
title = {E-Research Institutional Cloud Architecture (ERICA): An Orchestration Meta-Framework for Establishing Trusted Research Environments Using Public Cloud Computing.},
journal = {International journal of population data science},
volume = {6},
number = {1},
pages = {3373},
pmid = {42211407},
issn = {2399-4908},
mesh = {*Cloud Computing ; *Computer Security ; Internet ; Academia ; New South Wales ; *Research ; Confidentiality ; Australia ; Humans ; },
abstract = {INTRODUCTION: The E-Research Institutional Cloud Architecture (ERICA) is a code-driven orchestration framework that automates the configuration and management of Amazon Web Services (AWS) resources to provide trusted research environments (TREs) for sensitive data. Independent ERICA TREs are now operational in universities and government agencies. The framework was developed by the University of New South Wales, Australia, with support from the Australian Research Data Commons.

OBJECTIVES: ERICA was designed to overcome the limitations of traditional on-premise TREs by providing secure, scalable, and flexible cloud environments that protect privacy while enabling advanced, data-intensive research.

APPROACH: Using an infrastructure-as-code model, ERICA delivers consistent, reliable, and error-free setup of Project Spaces. It integrates robust security features, including encryption-at-rest and in transit, and multi-factor authentication. Hosted in AWS onshore data centres, ERICA ensures data sovereignty while supporting diverse operating systems and high-performance computing configurations. The architecture also allows rapid deployment of new AWS services, including generative AI tools, within research workspaces.

DISCUSSION: ERICA implements the Five Safes framework-covering safe projects, people, data, settings, and outputs-to ensure compliance and secure research. Its modular architecture enables multiple independent TREs, each governed by host-institution policies and capable of supporting hundreds of Project Spaces. This flexibility allows replication across any jurisdiction with AWS public cloud infrastructure. However, reliance on AWS introduces challenges, including charges in US dollars and delayed rollout of new services in smaller regions.

CONCLUSIONS: ERICA represents a step change in providing privacy-by-design cloud infrastructure for sensitive, data-intensive research. By combining strong governance with the scalability of AWS, it enables researchers to work securely with large, complex datasets while rapidly adopting cutting-edge analytical tools. ERICA TREs offer a replicable, future-proof model for supporting secure research at scale.},
}

*Cloud Computing
*Computer Security
Internet
Academia
New South Wales
*Research
Confidentiality
Australia
Humans

@article {pmid42214073,
year = {2026},
author = {Sparrock, LS and Vidrine, JI and Vinci, CE and El Naqa, IM and Sutton, SK and Salloum, RG and Dallery, J and Crane, TE and Penedo, FJ and Brockway, SJ and Jones, SR and Hoogland, CE and Reich, RR and Gonzalez-Calderon, G and Simmons, VN and Vidrine, DJ},
title = {Creation of an mHealth Infrastructure to Support the Development and Delivery of mHealth Interventions: Protocol for Demonstration Projects Addressing Smoking Cessation in Cancer Care.},
journal = {JMIR research protocols},
volume = {15},
number = {},
pages = {e92288},
doi = {10.2196/92288},
pmid = {42214073},
issn = {1929-0748},
mesh = {Humans ; *Smoking Cessation/methods ; *Telemedicine ; *Neoplasms/therapy ; Mobile Applications ; Digital Health ; },
abstract = {BACKGROUND: Cancer remains a leading cause of morbidity worldwide. To reduce this burden, scalable, effective approaches are needed to address modifiable risk factors for cancer and support behavioral self-management. With smartphone ownership now nearly ubiquitous, mobile health (mHealth) interventions offer a powerful means to extend the reach, accessibility, and sustainability of evidence-based treatments for a variety of modifiable risk factors (eg, excessive alcohol use, physical inactivity, poor diet, and smoking). Moreover, the flexibility of mHealth platforms enables efficient delivery of novel interventions, supports innovative study designs, and facilitates real-time data collection to advance public health research.

OBJECTIVE: Despite the great potential of mHealth interventions, developing high-quality mHealth tools is complex, time-consuming, and resource-intensive. To address these challenges, we are developing a coordinated, accessible, research-grade infrastructure for mHealth app development, testing, and dissemination.

METHODS: The mHealth Florida infrastructure (mFLi) will provide a comprehensive, low-code software platform that enables researchers to build apps compatible with major mobile operating systems, namely, Apple iOS and Google Android. Through a modular interface, users will select from a menu of prebuilt features to tailor functionality to specific study needs. The platform will include 3 integrated environments (development, testing, and production), allowing researchers to prototype, evaluate, and deploy mHealth interventions. This infrastructure will be developed and maintained by a multidisciplinary team, ensuring that the platform is technically robust and usable and adheres to institutional and regulatory standards. To demonstrate the platform's functionality, utility, and adaptability, a multisite study comprising three initial projects focused on smoking cessation among patients with cancer is being conducted: (1) participant screening and enrollment, (2) randomization and treatment delivery, and (3) data processing using machine learning methods with on-device and cloud-based approaches.

RESULTS: This study was funded in May 2023, and ethics approval was obtained from all involved sites' institutional review boards between February 2024 and October 2025. Recruitment began in March 2025 and enrollment is ongoing. As of January 2026, 41% (37/90) of the target sample have been enrolled and 21% (19/90) have completed their 6-month assessment. Data collection will be completed once the final participant completes their 6-month assessment (expected May 2027), with analyses commencing thereafter. Study findings are anticipated to be published in a peer-reviewed journal in 2027.

CONCLUSIONS: Collectively, these projects will illustrate how mFLi can streamline app development, facilitate rapid translation of research into practice, and reduce barriers for researchers and developers. Ultimately, mFLi is designed to accelerate innovation in mHealth research, enhance access to behavioral interventions, and improve health outcomes among diverse populations.

TRIAL REGISTRATION: ClinicalTrials.gov NCT06909357; https://clinicaltrials.gov/study/NCT06909357.

PRR1-10.2196/92288.},
}

Humans
*Smoking Cessation/methods
*Telemedicine
*Neoplasms/therapy
Mobile Applications
Digital Health

@article {pmid42215516,
year = {2026},
author = {Li, X and Li, J and Xing, Z and Liu, S},
title = {Monitoring alpine wetland of Haizishan using Otsu method and Sentinel-1 SAR in Hengduan Mountains, China.},
journal = {Scientific reports},
volume = {},
number = {},
pages = {},
doi = {10.1038/s41598-026-46543-5},
pmid = {42215516},
issn = {2045-2322},
abstract = {This study focuses on delineating water extent within the alpine wetlands of Haizishan, Qinghai-Tibet Plateau, a meteorologically complex region. Situated within a 6,000 km[2] Quaternary ice sheet landscape with over 600 water bodies, this area offers a crucial setting to examine climate change impacts on high-altitude wetland systems. Employing Sentinel-1 SAR data and the Google Earth Engine (GEE) platform, we implemented an Improved OTSU algorithm, achieving > 95% accuracy in water surface delineation despite persistent cloud cover. Notably, wetland water extent expanded from 52.9 km[2] to 54.8 km[2] between 2015 and 2022, correlating with increased precipitation. This expansion, and its precipitation-driven nature, distinguishes these non-glacial-fed wetlands hydrologically from glacial water bodies in the broader region, particularly impacting shallower wetland areas sensitive to thermal variations.},
}

@article {pmid42215677,
year = {2026},
author = {Rout, J and Mishra, M and Barik, RC and Yadav, DK and Prasad, DK and Sekh, AA},
title = {Stacked multi-fusion CNN: an adaptive attention model for privacy preserving deepfake forensics.},
journal = {Scientific reports},
volume = {},
number = {},
pages = {},
doi = {10.1038/s41598-026-55325-y},
pmid = {42215677},
issn = {2045-2322},
abstract = {The emergence of Generative Artificial Intelligence (Gen-AI) and Generative Adversarial Network (GAN)-based deepfakes poses significant security risks in sociocultural and sociopolitical domains. This necessitates the development of advanced and effective detection methods to prevent vulnerability in social networks. Classical Machine Learning (ML) algorithms have their limitations, especially in classifying the deepfakes. To address these issues, this paper suggests a privacy-preserving Stacked Multi-Fusion (SMF) Convolutional Neural Network (CNN) approach to classify deepfakes. An improved CNN model is proposed, integrating an adaptive multi-scale attention framework with enhanced residual blocks and a Squeeze-and-Excitation (SE) mechanism. The selection of these components is backed with an ablation study to report the individual contribution to the overall optimal architecture. A hybrid lossless multilayer cryptosystem based on a chaos-based approach, Deoxyribonucleic Acid (DNA)-based computing, etc., is developed to secure images in cloud storage. The efficacy of the proposed SMF model is validated using the 140K Real and Fake Faces (RFF) image dataset. The proposed approach was found to achieve stable performance with minimal variation in various tests. It achieved a test accuracy and ROC-AUC of 97.80 and 99.79, respectively. This paper provides comprehensive relevant factors for building effective synthetic media detection systems.},
}

@article {pmid42206982,
year = {2026},
author = {Spichak, S},
title = {Biocomputing: Beyond the Hype.},
journal = {Journal of medical Internet research},
volume = {28},
number = {},
pages = {e100949},
doi = {10.2196/100949},
pmid = {42206982},
issn = {1438-8871},
mesh = {Humans ; *Computational Biology ; Artificial Intelligence ; },
abstract = {Biocomputing is a nascent but rapidly developing field at the intersection of biology and computer science. In this News and Perspectives article, JMIR Correspondent Simon Spichak reports on its current and potential applications for health care research and beyond.},
}

Humans
*Computational Biology
Artificial Intelligence

@article {pmid42194316,
year = {2026},
author = {Oloko-Oba, M and Esenogho, E and Aruleba, K},
title = {From Biosignals to Bedside: A Review of Real-Time Edge Machine Learning for Wearable Health Monitoring.},
journal = {Bioengineering (Basel, Switzerland)},
volume = {13},
number = {5},
pages = {},
doi = {10.3390/bioengineering13050559},
pmid = {42194316},
issn = {2306-5354},
abstract = {Wearable devices increasingly capture biosignals such as electrocardiograms, photoplethysmograms, inertial signals, and electrodermal activity during daily life, enabling earlier detection and continuous monitoring outside the clinic. Real-time edge machine learning can convert these streams into timely, privacy-preserving inference by placing computation on a wearable (device-only) or a paired phone, with intermittent cloud assist used selectively for dashboards, summarisation, and lifecycle management. Clinical adoption remains uneven because free-living data are noisy, labels are often delayed, and device ecosystems evolve over time. This narrative review organises the literature as an end-to-end deployment pathway: sensing and artefact management, streaming windowing and multimodal alignment, and model families suited to on-device inference. We compare classical feature-based pipelines with learned representations, including compact CNN/TCN and recurrent and efficient attention-based models, and discuss when self-supervised pretraining and distillation are most useful in low-label settings. We then synthesise deployment engineering levers (quantisation, pruning, and distillation) and benchmarking requirements, emphasising runtime constraints that determine feasibility: latency per update, peak RAM, energy per inference, duty cycle, and thermal behaviour. Applications are grouped across cardiovascular monitoring, blood pressure and haemodynamics, sleep and respiration, and movement and stress, with explicit attention to false-alert burden, adherence, and workflow integration. To support translation, we provide a validation ladder and a reliability toolkit covering calibration, uncertainty-aware thresholds and deferral, drift monitoring triggers, and safe update governance. The novelty of this review is a deployment-oriented synthesis that ties modelling choices to edge tiers and resource budgets and provides reusable reporting templates, including an edge-cost card and comparative tables spanning modalities, models, deployment levers, applications, and reliability requirements.},
}

@article {pmid42195503,
year = {2026},
author = {Zhang, Z},
title = {A Review of Embedded Artificial Intelligence Research (2023-2026): Technological Advancements, Representative Advances, and Future Prospects.},
journal = {Micromachines},
volume = {17},
number = {5},
pages = {},
doi = {10.3390/mi17050586},
pmid = {42195503},
issn = {2072-666X},
abstract = {Since the publication of the "Review of Embedded Artificial Intelligence Research" in 2023, driven by innovations in hardware architectures, advances in lightweight algorithms, and the maturation of edge-cloud collaboration technologies, embedded artificial intelligence (embedded AI) has progressed from "technically feasible" to "large-scale deployment". As a continuation of that review, this article systematically surveys the core advances in embedded AI from 2023 to 2026. At the hardware level, it examines engineering progress in non-von Neumann architectures such as compute-in-memory and neuromorphic chips, as well as heterogeneous integration technologies. At the algorithmic level, it covers dynamic adaptive lightweighting, specialized edge-side optimization of large models (including on-device large language model fine-tuning and edge diffusion models), and lightweight multimodal approaches. In terms of deployment paradigms, it discusses edge-side full training, federated edge learning, edge-cloud collaborative intelligence, and emerging paradigms. At the application level, it illustrates the "perception-decision-execution" pipeline in industrial IoT, wearable healthcare, autonomous driving, embodied intelligence, and smart agriculture. The article also analyzes core challenges including ultra-low-power design for extreme scenarios, cross-platform standardization, edge-side data security and privacy, and model robustness in complex environments. Based on these findings, four research directions are proposed to guide future work.},
}

@article {pmid42197891,
year = {2026},
author = {Rosa-Bilbao, J},
title = {A Low-Code Containerized Edge Architecture for IIoT Telemetry Orchestration: Mitigating Cloud API Rate Limits Through Dual-Path Routing.},
journal = {Sensors (Basel, Switzerland)},
volume = {26},
number = {10},
pages = {},
doi = {10.3390/s26103082},
pmid = {42197891},
issn = {1424-8220},
support = {PID2021-122215NB-C33//Ministerio de Ciencia e Innovación/ ; },
abstract = {This paper investigates whether a low-code workflow engine can operate as practical Industrial Internet of Things (IIoT) middleware at the edge when cloud application programming interface (API) rate limits make direct telemetry upload unsustainable. The main contribution is a dual-path architecture in which a Hot Path persists all telemetry locally, while a Cold Path selectively forwards only anomalous or summary events to cloud services. The architecture is implemented as a lightweight containerized stack based on n8n, Eclipse Mosquitto, InfluxDB, and Grafana, and evaluated on a Raspberry Pi 4 under baseline, cloud-only saturation, and edge-filtered stress scenarios. Under the cloud-only condition, the external endpoint is throttled to approximately 60 requests/min, yielding a rejection rate of 98.0% (95% Wilson confidence interval: 97.43-98.44%). Under the dual-path condition, the same inbound load is fully retained locally while outbound cloud traffic is reduced by 98.0%, thereby avoiding throttling without sacrificing edge-side data fidelity. The measured Hot Path processing latency remains around 5 ms on average, with observed peaks below 10 ms, which is compatible with soft real-time monitoring workloads. Compared with more established low-code tools such as Node-RED, the novelty of the study is not the existence of visual orchestration itself, but the combination of containerized deployment, explicit hot/cold decoupling, and an empirical rate-limit mitigation analysis focused on low-cost edge hardware.},
}

@article {pmid42197916,
year = {2026},
author = {Chen, GH and Ma, HY and Yu, W and Wen, J and Chen, K and Wang, JJ and Chen, SD and Sun, YL},
title = {Terminal-Edge-Cloud Collaborative Computation Offloading and Resource Allocation Strategy Based on Improved Mayfly Algorithm for District Heating Systems.},
journal = {Sensors (Basel, Switzerland)},
volume = {26},
number = {10},
pages = {},
doi = {10.3390/s26103110},
pmid = {42197916},
issn = {1424-8220},
abstract = {The rapid digitalization of district heating systems (DHSs) has driven the large-scale deployment of thermal Internet of Things (TIoT) sensors, which generate massive real-time operational data. Traditional centralized computing architectures struggle to process massive concurrent data. Furthermore, they fail to balance the stringent low-latency demands of real-time control tasks with the low-energy constraints of battery-powered terminal devices. To solve the complex problem of minimizing the weighted sum of system latency and energy consumption, we propose an Improved Mayfly Algorithm (IMA). The algorithm integrates five targeted structural enhancements: random position update masking, differential evolution (DE)-based crossover, targeted subset mutation with boundary scaling, adaptive population reset mechanism, and simulated annealing (SA)-driven local search, to efficiently navigate the high-dimensional rugged decision space and mitigate premature convergence. Extensive simulation results show that the proposed collaborative architecture achieves the lowest total system cost compared with traditional isolated computing paradigms (local-only, edge-only, and cloud-only). Notably, the proposed IMA reduces the total baseline weighted cost by 17.2% compared with the standard MA. Furthermore, under maximum practical industrial workloads (750 concurrent tasks, representing a highly complex 2250-dimensional MINLP space), the IMA maintains strong scalability and dominance, outperforming the second-best algorithm (BWO) by 15.8%. This research provides a low-latency, energy-efficient scheduling solution for TIoT-enabled DHS, and offers technical support for the intelligent and low-carbon transformation of urban energy infrastructure.},
}

@article {pmid42198259,
year = {2026},
author = {Altharawi, A and Alqahtani, SM},
title = {Integrative Computational Chemistry Approaches in Modern Drug Discovery: Advances in Docking, Pharmacophore Modeling, Molecular Dynamics, and Virtual Screening.},
journal = {Pharmaceutics},
volume = {18},
number = {5},
pages = {},
doi = {10.3390/pharmaceutics18050565},
pmid = {42198259},
issn = {1999-4923},
abstract = {Computational chemistry has played a central role in early-stage drug discovery by accelerating target selection, hit identification, and lead optimization. This review summarizes recent developments in molecular docking, pharmacophore modeling, molecular dynamics (MD), and virtual screening (VS), with a focus on their application in practical drug discovery workflows. Advances in docking protocols, including consensus scoring, physics-based rescoring, and ensemble approaches, addressed the challenges of receptor flexibility. Both ligand-based and structure-based pharmacophore models facilitated scaffold hopping and guided library prioritization. MD simulations were used to assess binding pose stability, identify cryptic binding pockets, and characterize solvent interactions. These simulations also supported free-energy calculations using endpoint and alchemical methods. Large-scale VS campaigns employed curated compound libraries, often composed of make-on-demand molecules, and relied on high-performance computing or cloud infrastructure to screen up to 10[9] compounds. Hits were validated using orthogonal biophysical assays and filtered by absorption, distribution, metabolism, excretion, and toxicity (ADMET) predictions. Integrated pipelines combining pharmacophore modeling, docking, MD, and free-energy calculations improved enrichment rates and reduced the number of compounds requiring synthesis. Several case studies demonstrated the identification of nanomolar-affinity leads from ultra-large screening campaigns. The review also addressed ongoing challenges, such as inconsistent scoring of binding affinity, protonation, and tautomeric errors, dataset bias, and reproducibility issues. Strategies to mitigate these limitations included standardized library preparation, adherence to FAIR (Findable, Accessible, Interoperable, and Reusable) data principles, and the use of prospective benchmarking protocols. The review discussed emerging trends, including the use of quantum chemistry for electronic structure refinement, ensemble docking guided by cryo-electron microscopy (cryo-EM) data, and the integration of computational tools with automated synthesis and high-throughput screening in closed-loop discovery systems. These approaches have the potential to accelerate the design-make-test cycle, increase hit novelty, and improve decision-making in early drug development programs.},
}

@article {pmid42199228,
year = {2026},
author = {Ye, Z and Liao, L and Qiu, G and Xiao, Y and Huang, D and Duan, Z and Pu, S and Wen, J},
title = {Application of crop growth models in crop yield assessment.},
journal = {Frontiers in plant science},
volume = {17},
number = {},
pages = {1819890},
pmid = {42199228},
issn = {1664-462X},
abstract = {Global food security is facing formidable challenges due to rising temperatures, frequent extreme weather events, growing water scarcity, cropland reduction, fluctuations in international food trade, and rising food demand. Crop production systems are complex, multi-factor dynamic systems influenced collectively by crop cultivars, climatic conditions, soil properties, and management practices, which exhibit strong spatiotemporal variability. Crop growth models have emerged as essential tools in smart agriculture, integrating knowledge from crop physiology, ecology, meteorology, soil science, and agronomy to simulate crop growth processes dynamically. This systematic review focuses on six key aspects of crop growth modeling: (1) introduction of major crop models; (2) assessing climate change impacts on crop yields; (3) predicting yield potential and yield gaps; (4) identifying yield-limiting factors; (5) formulating adaptation strategies; and (6) challenges and future research directions. Future research should focus on the deep integration of crop growth models with remote sensing, the Internet of Things (IoT), big data, cloud computing, and artificial intelligence technologies to establish intelligent "space-air-ground" decision-making systems that support precision, unmanned, and climate-resilient agriculture.},
}

@article {pmid42200970,
year = {2026},
author = {Valsalan, P and Siddiqui, MM},
title = {AI-Driven Hybrid Detection and Classification Framework for Secure Sleep Health IoT Networks.},
journal = {Clocks & sleep},
volume = {8},
number = {2},
pages = {},
doi = {10.3390/clockssleep8020023},
pmid = {42200970},
issn = {2624-5175},
abstract = {Sleep disorders, such as insomnia, obstructive sleep apnea (OSA), narcolepsy, REM sleep behavior disorder, and circadian rhythm disturbances, represent a rapidly expanding global health burden that is strongly associated with cardiovascular, metabolic, neurological, and psychiatric diseases. Advancements in wearable sensing technologies and Internet of Medical Things (IoMT) infrastructures have expanded the possibilities for continuous, home-based sleep assessment beyond conventional polysomnography laboratories. These Sleep Health Internet of Things (S-HIoT) systems combine multimodal physiological sensing (EEG, ECG, SpO2, respiratory effort and actigraphy) with wireless communication and cloud-based analytics for automated sleep-stage classification and disorder detection. Nonetheless, the digitization of sleep medicine brings about significant cybersecurity concerns. The constant transmission of sensitive biomedical information makes S-HIoT networks open to anomalous traffic flows, signal manipulation, replay attacks, spoofing, and data integrity violation. Existing studies mostly focus on analyzing physiological signals and network intrusion detection independently, resulting in a systemic vulnerability of cyber-physical sleep monitoring ecosystems. With the aim of addressing this empirical deficiency, this review integrates emerging advances (2022-2026) in the AI-assisted categorization of sleep phases and IoMT anomaly detector designs on the finer analysis of CNN, LSTM/BiLSTM, Transformer-based systems, and a component part of federated schemes and the lightweight, edge-deployable intruder assessor models available. The aim of this study is to uncover a gap in the literature: integrated architectures to trade off audiences of faithfulness of physiological modeling with communication-layer security. To counter it, we present a single framework to include CNN-based spatial feature extraction, Bidirectional Long Short-Term Memory (BiLSTM)-based temporal models and Random Forest-based ensemble classification using a dual task-learning approach. We propose a multi-objective optimization framework to jointly optimize the performance of sleep-stage prediction and that of network anomaly detection. Performance on publicly available datasets (Sleep-EDF and CICIoMT2024) confirms that hybrid integration can be tailored to achieve high accuracy [99.8% sleep staging; 98.6% anomaly detection] whilst being characterized by low inference latency (<45 ms), which is promising for feasibility in real-time deployment in view of targeting edge devices. This work presents a comprehensive framework for developing secure, intelligent, and clinically robust digital sleep health ecosystems by bridging chronobiological signal modeling with cybersecurity mechanisms. Furthermore, it highlights future research directions, including explainable AI, federated secure learning, adversarial robustness, and energy-aware edge optimization.},
}

@article {pmid42201543,
year = {2026},
author = {Huynh, T and Nguyen, M and Ha, HTT and Ngo, L},
title = {Development of a serverless, interactive application for Alzheimer's disease detection and visualization using MRI images.},
journal = {The neuroradiology journal},
volume = {},
number = {},
pages = {19714009261456608},
doi = {10.1177/19714009261456608},
pmid = {42201543},
issn = {2385-1996},
abstract = {BackgroundAlzheimer's disease requires early detection for effective intervention with disease-modifying treatments, yet significant implementation barriers persist in clinical practice, including limited computational infrastructure and the gap between research model performance and practical deployment in resource-constrained settings.ObjectivesTo develop and evaluate a computer-aided diagnosis system for classifying cognitive states (AD, MCI, and CN) from structural MRI, with automated preprocessing and cost-effective cloud deployment suitable for resource-constrained healthcare facilities.DesignComputer-aided diagnosis system integrating neural architecture search with serverless cloud infrastructure.MethodsA multi-view MRI analysis model was optimized through neural architecture search, incorporating Universal Inverted Bottleneck blocks and Kolmogorov-Arnold Networks. Automated MRI preprocessing using FSL was deployed through cloud-based serverless functions for scalable image processing. Evaluation used the ADNI dataset (1687 individuals: 368 AD, 625 MCI, and 694 CN). A web application was developed providing patient management, MRI visualization, and automated diagnostic prediction.ResultsThe model achieved 86.7% accuracy and 0.900 AUC in three-class classification with 1.7 million parameters. High specificity was observed across all classes (CN: 91.0%, MCI: 91.8%, AD: 97.3%), with 100% CN-AD specificity ensuring no AD cases were misclassified as cognitively normal. Operational costs were approximately 0.028 USD per diagnosis for typical hospital workloads.ConclusionThis system provides a cost-effective approach for early Alzheimer's diagnosis accessible to resource-constrained environments. Despite challenges in MCI classification, the combination of neural architecture search with serverless deployment demonstrates progress toward clinically deployable automated AD detection. Future work should focus on prospective clinical validation and integration of interpretability features.},
}

@article {pmid42202663,
year = {2026},
author = {Pescol, F and Buonocore, TM and Tibollo, V and Failla, G and Traversi, E and La Rovere, MT and Sacchi, L and Ricciardi, W and Bellazzi, R},
title = {Evaluating large language models for structuring cardiology reports: a real-world clinical study on patient subtyping and trial recruitment.},
journal = {International journal of medical informatics},
volume = {217},
number = {},
pages = {106509},
doi = {10.1016/j.ijmedinf.2026.106509},
pmid = {42202663},
issn = {1872-8243},
abstract = {BACKGROUND: Artificial Intelligence (AI) methods have emerged as useful tools for supporting patient recruitment in clinical trials (CTs). Despite several studies having recently proposed promising applications of Large Language Model (LLMs) for patient recruitment in CTs, their implementation in routine clinical practice remains limited.

METHODS: In this study, we present a comprehensive pipeline, developed and tested in a real-world clinical setting, to obtain highly detailed patient subtyping and eligibility assessment for specific CTs. Our solution leverages cardiological discharge letters, a rich yet underutilized source of patient data, to extract detailed structured clinical information through LLMs. Patient subtyping and eligibility assessment are performed through a rule-based approach, based on the extracted information, to maximize deterministic and interpretable outputs. We employed OpenAI's GPT-4.1 within the cloud-based service Microsoft Azure Machine Learning Studio, deployed in the hospital infrastructure. Validation was conducted on a sample of 100 discharge letters through exact-match comparison between the model's output and a ground-truth template, pre-populated by expert clinicians.

RESULTS: Our results confirm the feasibility and effectiveness of the proposed approach in real-world clinical scenarios. GPT-4.1 achieved high values of information extraction accuracy for most clinical variables (0.94 ± 0.08), resulting in a limited number of false negatives (FN) and false positives (FP) in both patient subtyping (0.12 and 0.13, respectively) and eligibility assessment. At the criterion-level, the proportion of FNs and FPs was below 3% for most criteria (13 and 11 of the 14 criteria examined, respectively).

CONCLUSION: Overall, our study presents a notable step towards the integration of AI-driven approaches into real-world clinical practice for patient recruitment in CTs, highlighting both its practicality and effectiveness in meeting the stringent demands of healthcare settings.},
}