A Systematic Review of Federated Learning in the Healthcare Area: From the Perspective of Data Properties and Applications

Applied Sciences, Journal Year: 2021, Volume and Issue: 11(23), P. 11191 - 11191

Published: Nov. 25, 2021

Recent advances in deep learning have shown many successful stories smart healthcare applications with data-driven insight into improving clinical institutions’ quality of care. Excellent models are heavily data-driven. The more data trained, the robust and generalizable performance model. However, pooling medical centralized storage to train a model faces privacy, ownership, strict regulation challenges. Federated resolves previous challenges shared global using central aggregator server. At same time, patient remain local party, maintaining anonymity security. In this study, first, we provide comprehensive, up-to-date review research employing federated applications. Second, evaluate set recent from data-centric perspective learning, such as partitioning characteristics, distributions, protection mechanisms, benchmark datasets. Finally, point out several potential future directions

Language: Английский

---

Federated Learning for Medical Applications: A Taxonomy, Current Trends, Challenges, and Future Research Directions

IEEE Internet of Things Journal, Journal Year: 2023, Volume and Issue: 11(5), P. 7374 - 7398

Published: Nov. 1, 2023

With the advent of Internet Things (IoT), artificial intelligence (AI), machine learning (ML), and deep (DL) algorithms, landscape data-driven medical applications has emerged as a promising avenue for designing robust scalable diagnostic prognostic models from data. This gained lot attention both academia industry, leading to significant improvements in healthcare quality. However, adoption AI-driven still faces tough challenges, including meeting security, privacy, Quality-of-Service (QoS) standards. Recent developments federated (FL) have made it possible train complex machine-learned distributed manner become an active research domain, particularly processing data at edge network decentralized way preserve privacy address security concerns. To this end, article, we explore present future FL technology where sharing is challenge. We delve into current trends their outcomes, unraveling complexities reliable models. article outlines fundamental statistical issues FL, tackles device-related problems, addresses navigates complexity concerns, all while highlighting its transformative potential field. Our study primarily focuses on context global cancer diagnosis. highlight enable computer-aided diagnosis tools that challenge with greater effectiveness than traditional methods. literature shown are generalize well new data, which essential applications. hope comprehensive review will serve checkpoint field, summarizing state art identifying open problems directions.

Language: Английский

---

Federated learning for secure IoMT-applications in smart healthcare systems: A comprehensive review

Knowledge-Based Systems, Journal Year: 2023, Volume and Issue: 274, P. 110658 - 110658

Published: May 22, 2023

Recent developments in the Internet of Things (IoT) and various communication technologies have reshaped numerous application areas. Nowadays, IoT is assimilated into medical devices equipment, leading to progression Medical (IoMT). Therefore, IoMT-based healthcare applications are deployed used day-to-day scenario. Traditionally, machine learning (ML) models use centralized data compilation that impractical pragmatic frameworks due rising privacy security issues. Federated Learning (FL) has been observed as a developing distributed collective paradigm, most appropriate for modern framework, manages stakeholders (e.g., patients, hospitals, laboratories, etc.) carry out training without actual exchange sensitive data. Consequently, this work, authors present an exhaustive survey on FL-based IoMT smart frameworks. First, introduced devices, their types, applications, datasets, framework detail. Subsequently, concept FL, its domains, tools develop FL discussed. The significant contribution deploying secure systems presented by focusing patents, real-world projects, datasets. A comparison techniques with other schemes ecosystem also presented. Finally, discussed challenges faced potential future research recommendations deploy

Language: Английский

---

Federated Learning for Healthcare Domain - Pipeline, Applications and Challenges

ACM Transactions on Computing for Healthcare, Journal Year: 2022, Volume and Issue: 3(4), P. 1 - 36

Published: May 12, 2022

Federated learning is the process of developing machine models over datasets distributed across data centers such as hospitals, clinical research labs, and mobile devices while preventing leakage. This survey examines previous studies on federated in healthcare sector a range use cases applications. Our shows what challenges, methods, applications practitioner should be aware topic learning. paper aims to lay out existing list possibilities for industries.

Language: Английский

---

Hybrid Classifier-Based Federated Learning in Health Service Providers for Cardiovascular Disease Prediction

Applied Sciences, Journal Year: 2023, Volume and Issue: 13(3), P. 1911 - 1911

Published: Feb. 1, 2023

One of the deadliest diseases, heart disease, claims millions lives every year worldwide. The biomedical data collected by health service providers (HSPs) contain private information about patient and are subject to general privacy concerns, sharing is restricted under global laws. Furthermore, collection have a significant network communication cost lead delayed disease prediction. To address training latency, cost, single point failure, we propose hybrid framework at client end HSP consisting modified artificial bee colony optimization with support vector machine (MABC-SVM) for optimal feature selection classification disease. For server, proposed federated matched averaging overcome issues in this paper. We tested evaluated our technique compared it standard learning techniques on combined cardiovascular dataset. Our experimental results show that improves prediction accuracy 1.5%, achieves 1.6% lesser error, utilizes 17.7% rounds reach maximum accuracy.

Language: Английский

---

A Systematic Review on Federated Learning in Medical Image Analysis

IEEE Access, Journal Year: 2023, Volume and Issue: 11, P. 28628 - 28644

Published: Jan. 1, 2023

Federated Learning (FL) obtained a lot of attention to the academic and industrial stakeholders from beginning its invention. The eye-catching feature FL is handling data in decentralized manner which creates privacy preserving environment Artificial Intelligence (AI) applications. As we know medical includes marginal private information patients demands excessive protection disclosure unexpected destinations. In this paper, performed Systematic Literature Review (SLR) published research articles on based image analysis. Firstly, have collected different databases followed by PRISMA guidelines, then synthesized selected articles, finally provided comprehensive overview topic. order do that extracted core associated with implementation imaging articles. our findings briefly presented characteristics federated models, performance achieved models exclusively results comparison traditional ML models. addition, discussed open issues challenges implementing mentioned recommendations for future direction particular field. We believe SLR has successfully summarized state-of-the-art methods analysis using deep learning.

Language: Английский

---

A comprehensive review on federated learning based models for healthcare applications

Artificial Intelligence in Medicine, Journal Year: 2023, Volume and Issue: 146, P. 102691 - 102691

Published: Oct. 30, 2023

Language: Английский

---

SAM: Self-augmentation mechanism for COVID-19 detection using chest X-ray images

Knowledge-Based Systems, Journal Year: 2022, Volume and Issue: 241, P. 108207 - 108207

Published: Jan. 17, 2022

COVID-19 is a rapidly spreading viral disease and has affected over 100 countries worldwide. The numbers of casualties cases infection have escalated particularly in with weakened healthcare systems. Recently, reverse transcription-polymerase chain reaction (RT-PCR) the test choice for diagnosing COVID-19. However, current evidence suggests that infected patients are mostly stimulated from lung after coming contact this virus. Therefore, chest X-ray (i.e., radiography) CT can be surrogate some where PCR not readily available. This forced scientific community to detect images recently proposed machine learning methods offer great promise fast accurate detection. Deep convolutional neural networks (CNNs) been successfully applied radiological imaging improving accuracy diagnosis. performance remains limited due lack representative available public benchmark datasets. To alleviate issue, we propose self-augmentation mechanism data augmentation feature space rather than using reconstruction independent component analysis (RICA). Specifically, unified architecture which contains deep network (CNN), mechanism, bidirectional LSTM (BiLSTM). CNN provides high-level features extracted at pooling layer chooses most relevant generates low-dimensional augmented features. Finally, BiLSTM used classify processed sequential information. We conducted experiments on three publicly databases show approach achieves state-of-the-art results 97%, 84% 98%. Explainability carried out visualization through PCA projection t-SNE plots.

Language: Английский

---

FedEHR: A Federated Learning Approach towards the Prediction of Heart Diseases in IoT-Based Electronic Health Records

Diagnostics, Journal Year: 2023, Volume and Issue: 13(20), P. 3166 - 3166

Published: Oct. 10, 2023

In contemporary healthcare, the prediction and identification of cardiac diseases is crucial. By leveraging capabilities Internet Things (IoT)-enabled devices Electronic Health Records (EHRs), healthcare sector can largely benefit to improve patient outcomes by increasing accuracy disease prediction. However, protecting data privacy essential promote participation adhere rules. The suggested methodology combines EHRs with IoT-generated health predict heart disease. For its capacity manage high-dimensional choose pertinent features, a soft-margin L1-regularised Support Vector Machine (sSVM) classifier used. large-scale sSVM problem successfully solved using cluster primal–dual splitting algorithm, which improves computational complexity scalability. integration federated learning provides cooperative predictive analytics that upholds privacy. use framework in this study, focus on peer-to-peer applications, crucial for enabling collaborative modeling while confidentiality each participant’s private medical information.

Language: Английский

---

The economics of deep and machine learning-based algorithms for COVID-19 prediction, detection, and diagnosis shaping the organizational management of hospitals

Oeconomia Copernicana, Journal Year: 2024, Volume and Issue: 15(1), P. 27 - 58

Published: March 30, 2024

Research background: Deep and machine learning-based algorithms can assist in COVID-19 image-based medical diagnosis symptom tracing, optimize intensive care unit admission, use clinical data to determine patient prioritization mortality risk, being pivotal qualitative provision, reducing errors, increasing survival rates, thus diminishing the massive healthcare system burden relation severe inpatient stay duration, while operational costs throughout organizational management of hospitals. Data-driven financial scenario-based contingency planning, predictive modelling tools, risk pooling mechanisms should be deployed for additional equipment unforeseen demand expenses. Purpose article: We show that deep decision making systems likelihood treatment outcomes with regard susceptible, infected, recovered individuals, performing accurate analyses by modeling based on vital signs, surveillance data, infection-related biomarkers, furthering hospital facility optimization terms bed allocation. Methods: The review software employed article screening quality evaluation were: AMSTAR, AXIS, DistillerSR, Eppi-Reviewer, MMAT, PICO Portal, Rayyan, ROBIS, SRDR. Findings & value added: support tools forecast spread, confirmed cases, infection rates data-driven appropriate resource allocations effective therapeutic protocol development, determining suitable measures regulations using symptoms comorbidities, laboratory records across units, impacting financing infrastructure. As a result heightened personal protective equipment, pharmacy medication, outpatient treatment, supplies, revenue loss vulnerability occur, also due expenses related hiring staff critical expenditures. Hospital care, screening, capacity expansion, lead further losses affecting frontline workers patients.

Language: Английский

---