Hierarchical Federated Learning With Social Context Clustering-Based Participant Selection for Internet of Medical Things Applications

IEEE Transactions on Computational Social Systems, Journal Year: 2023, Volume and Issue: 10(4), P. 1742 - 1751

Published: April 4, 2023

The proliferation in embedded and communication technologies made the concept of Internet Medical Things (IoMT) a reality. Individuals' physical physiological status can be constantly monitored, numerous data collected through wearable mobile devices. However, silo individual brings limitations to existing machine learning approaches correctly identify user's health status. Distributed paradigms, such as federated learning, offer potential solution for privacy-preserving knowledge sharing without sending raw personal data. is vulnerable harmful participants that degrade overall model quality by low-quality Therefore, it critical select suitable ensure accuracy efficiency learning. In this article, unique clustering-based approach proposed use social context participant selection. Different edge groups will established, group-specific performed. models various further aggregated strengthen robustness global model. experimental results demonstrated selection, hierarchical achieve better with less two different IoMT applications ECG human motion monitoring. This shows efficacy method improving performance applications.

Language: Английский

---

EEG-based emotion recognition using hybrid CNN and LSTM classification

Frontiers in Computational Neuroscience, Journal Year: 2022, Volume and Issue: 16

Published: Oct. 7, 2022

Emotions are a mental state that is accompanied by distinct physiologic rhythm, as well physical, behavioral, and changes. In the latest days, physiological activity has been used to study emotional reactions. This describes electroencephalography (EEG) signals, brain wave pattern, emotion analysis all of these interrelated based on consequences human behavior Post-Traumatic Stress Disorder (PTSD). Post-traumatic stress disorder effects for long-term illness associated with considerable suffering, impairment, social/emotional impairment. PTSD connected subcortical responses injury memories, thoughts, emotions alterations in circuitry. Predominantly EEG signals way examining electrical potential feelings cum expression every changing phenomenon individual faces. When going through literature there some lacunae while analyzing emotions. There exist reliability issues also masking real victims. Keeping this research gap hindrance faced previous researchers present aims fulfill requirements, efforts can be made overcome problem, proposed automated CNN-LSTM ResNet-152 algorithm. Compared existing techniques, techniques achieved higher level accuracy 98% applying hybrid deep learning

Language: Английский

---

Wearable and Flexible Sensor Devices: Recent Advances in Designs, Fabrication Methods, and Applications

Sensors, Journal Year: 2025, Volume and Issue: 25(5), P. 1377 - 1377

Published: Feb. 24, 2025

The development of wearable sensor devices brings significant benefits to patients by offering real-time healthcare via wireless body area networks (WBANs). These have gained traction due advantageous features, including their lightweight nature, comfortable feel, stretchability, flexibility, low power consumption, and cost-effectiveness. Wearable play a pivotal role in healthcare, defence, sports, health monitoring, disease detection, subject tracking. However, the irregular nature human poses challenge design such systems. This manuscript provides comprehensive review recent advancements flexible smart that can support next generation devices. Further, direct ink writing (DIW) (DW) methods has revolutionised new high-resolution integrated structures, enabling next-generation soft, flexible, stretchable Recognising importance keeping academia industry informed about cutting-edge technology time-efficient fabrication tools, this also thorough overview latest progress various for utilised WBAN evaluation using phantoms. An emerging challenges future research directions is discussed conclusion.

Language: Английский

---

Deep Collaborative Intelligence-Driven Traffic Forecasting in Green Internet of Vehicles

IEEE Transactions on Green Communications and Networking, Journal Year: 2022, Volume and Issue: 7(2), P. 1023 - 1035

Published: July 26, 2022

Accompanied with the development of green wireless communication, Internet Vehicles (GIoV) has been a latent solution for future transportation. Among them, intelligent traffic forecasting key nodes in GIoV is significant research topic. Much had devoted to this issue, and graph learning-based approaches seemed be promising solution. However, existing works concentrated more on graph-structured features yet neglected global reliability. To deal such work combines both deep embedding together proposes collaborative intelligence-driven model GIoV. By establishing reliable feature spaces flow prediction, efficiency expected promoted. Specifically, utilized generate abstract representation basic road networks, employed update different timestamps. Their collaboration contributes considerable In addition, experiments are also conducted real-world dataset, results indicate that deviation receives about 15%-25% reduction.

Language: Английский

---

A Flexible, Wearable, and Wireless Biosensor Patch with Internet of Medical Things Applications

Biosensors, Journal Year: 2022, Volume and Issue: 12(3), P. 139 - 139

Published: Feb. 22, 2022

Monitoring the vital signs and physiological responses of human body in daily activities is particularly useful for early diagnosis prevention cardiovascular diseases. Here, we proposed a wireless flexible biosensor patch continuous longitudinal monitoring different signals, including temperature, blood pressure (BP), electrocardiography. Moreover, these modalities tracking movement GPS locations emergency rescue have been included devices. We optimized design with high mechanical stretchability compatibility that can provide reliable long-term attachment to curved skin surface. Regarding smart healthcare applications, this research presents an Internet Things-connected platform consisting smartphone application, website service, database server, mobile gateway. The IoT has potential reduce demand medical resources enhance quality services. To further address advances non-invasive BP monitoring, deep learning architecture one-channel electrocardiogram signals introduced. performance estimation model was verified using independent dataset; experimental result satisfied Association Advancement Medical Instrumentation, British Hypertension Society standards results demonstrated practical application signal applications.

Language: Английский

---

A Vision-Based Framework for Predicting Multiple Sclerosis and Parkinson's Disease Gait Dysfunctions—A Deep Learning Approach

IEEE Journal of Biomedical and Health Informatics, Journal Year: 2022, Volume and Issue: 27(1), P. 190 - 201

Published: Sept. 20, 2022

This study examined the effectiveness of a vision-based framework for multiple sclerosis (MS) and Parkinson's disease (PD) gait dysfunction prediction. We collected video data from multi-view digital cameras during self-paced walking MS, PD patients age, weight, height gender-matched healthy older adults (HOA). then extracted characteristic 3D joint keypoints videos. In this work, we proposed data-driven methodology to classify strides in persons with MS (PwMS), (PwPD) HOA that may generalize across different tasks subjects. presented comprehensive quantitative comparison 16 diverse traditional machine deep learning (DL) algorithms. When generalizing comfortable (W) walking-while-talking (WT), multi-scale residual neural network achieved perfect accuracy AUC classifying individuals given disorder; subject generalization W trials, resulted highest 78.1% 0.87 (resp.), 1D convolutional (CNN) had 75% WT trials. Finally, when over new subjects tasks, again CNN top classification 79.3% 0.93 (resp.). work is first attempt apply demonstrate potential DL camera-based analysis neurological suggests viability inexpensive systems diagnosing certain disorders.

Language: Английский

---

MCNN: a multi-level CNN model for the classification of brain tumors in IoT-healthcare system

Journal of Ambient Intelligence and Humanized Computing, Journal Year: 2022, Volume and Issue: 14(5), P. 4695 - 4706

Published: Sept. 15, 2022

Language: Английский

---

A Privacy-Preserving Social Computing Framework for Health Management Using Federated Learning

IEEE Transactions on Computational Social Systems, Journal Year: 2022, Volume and Issue: 10(4), P. 1666 - 1678

Published: Dec. 22, 2022

Currently, health management driven by intelligent means is a general demand of social systems. Although number researchers have paid attention to such areas, they primarily focused on improving the performance algorithms. Such algorithms are mostly based central computing mode, where all user data aggregated together in cloud implement tasks. This poses great threat personal privacy due exposure outside world. To address this challenge, work uses federated learning mechanism and proposes privacy-preserving framework for management. User deposited different terminals prevent exposure. A group parameters pretrained each terminal an iteration then transferred center updating. After multiple rounds interactive training between terminals, recognition model finishes without direct access from other sources. Finally, also conducts experiments real-world dataset assess overall proposed approach.

Language: Английский

---

A Physics-Guided Deep Learning Approach for Functional Assessment of Cardiovascular Disease in IoT-Based Smart Health

IEEE Internet of Things Journal, Journal Year: 2023, Volume and Issue: 10(21), P. 18505 - 18516

Published: Jan. 30, 2023

The rapid development of the Internet Things (IoT) widely supports smart healthcare system. IoT-based health has significant importance for diagnosis cardiovascular disease (CVD) in clinical practice. Combined with advanced artificial intelligence techniques, provides valuable and accurate information remotely disease. functional assessment CVD is an essential task It aims to determine extent myocardial ischemia through measurement hemodynamic parameters coronary artery. However, adoption limited due potential risks high costs during measurements. Recent advances have enabled computation based on anatomical features arteries. existing methods still lack explainability prediction. To address this issue, we present a physics-guided deep learning network manner. We specifically design attentive effective by considering artery anatomy segments. obtain explainability, incorporate physical knowledge related blood flow into loss function. can ensure that follows laws. Extensive experiments are performed synthetic data set real-world set. results show our approach achieve physically consistent assessment. Moreover, method promotes deeper IoT field health.

Language: Английский

---

A systematic review of deep learning-based cervical cytology screening: from cell identification to whole slide image analysis

Artificial Intelligence Review, Journal Year: 2023, Volume and Issue: 56(S2), P. 2687 - 2758

Published: Oct. 5, 2023

Abstract Cervical cancer is one of the most common cancers in daily life. Early detection and diagnosis can effectively help facilitate subsequent clinical treatment management. With growing advancement artificial intelligence (AI) deep learning (DL) techniques, an increasing number computer-aided (CAD) methods based on have been applied cervical cytology screening. In this paper, we survey more than 80 publications since 2016 to provide a systematic comprehensive review DL-based First, concise summary medical biological knowledge pertaining cytology, hold firm belief that biomedical understanding significantly contribute development CAD systems. Then, collect wide range public datasets. Besides, image analysis approaches applications including cell identification, abnormal or area detection, region segmentation whole slide are summarized. Finally, discuss present obstacles promising directions for future research automated

Language: Английский

---