Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Federated Learning in Healthcare: Model Misconducts, Security, Challenges, Applications, and Future Research Directions -- A Systematic Review

Md Shahin Ali, Md Manjurul Ahsan, Lamia Tasnim, Sadia Afrin, Koushik Biswas, Md Maruf Hossain, Md Mahfuz Ahmed, Ronok Hashan, Md Khairul Islam, Shivakumar Raman

TL;DR

This paper surveys Federated Learning (FL) in healthcare, addressing privacy, security, and data-sharing challenges while cataloging model misconduct risks and threat models. It classifies FL variants (horizontal, vertical, transfer, domain adaptation, multitask, meta learning) and surveys healthcare applications (COVID-19 detection, EHR, medical imaging, breast-density analysis, monitoring), along with practical misconduct defenses and adversarial strategies. The study highlights non-IID data, device heterogeneity, and scalability as key obstacles, proposing future directions such as privacy-preserving techniques, fairness-aware optimization, and hybrid architectures (e.g., blockchain+FL) to enable robust, privacy-conscious deployment. The findings aim to guide researchers and practitioners in designing secure, scalable FL systems that improve clinical outcomes while safeguarding patient data and ensuring equitable access across diverse populations.

Abstract

Data privacy has become a major concern in healthcare due to the increasing digitization of medical records and data-driven medical research. Protecting sensitive patient information from breaches and unauthorized access is critical, as such incidents can have severe legal and ethical complications. Federated Learning (FL) addresses this concern by enabling multiple healthcare institutions to collaboratively learn from decentralized data without sharing it. FL's scope in healthcare covers areas such as disease prediction, treatment customization, and clinical trial research. However, implementing FL poses challenges, including model convergence in non-IID (independent and identically distributed) data environments, communication overhead, and managing multi-institutional collaborations. A systematic review of FL in healthcare is necessary to evaluate how effectively FL can provide privacy while maintaining the integrity and usability of medical data analysis. In this study, we analyze existing literature on FL applications in healthcare. We explore the current state of model security practices, identify prevalent challenges, and discuss practical applications and their implications. Additionally, the review highlights promising future research directions to refine FL implementations, enhance data security protocols, and expand FL's use to broader healthcare applications, which will benefit future researchers and practitioners.

Federated Learning in Healthcare: Model Misconducts, Security, Challenges, Applications, and Future Research Directions -- A Systematic Review

TL;DR

This paper surveys Federated Learning (FL) in healthcare, addressing privacy, security, and data-sharing challenges while cataloging model misconduct risks and threat models. It classifies FL variants (horizontal, vertical, transfer, domain adaptation, multitask, meta learning) and surveys healthcare applications (COVID-19 detection, EHR, medical imaging, breast-density analysis, monitoring), along with practical misconduct defenses and adversarial strategies. The study highlights non-IID data, device heterogeneity, and scalability as key obstacles, proposing future directions such as privacy-preserving techniques, fairness-aware optimization, and hybrid architectures (e.g., blockchain+FL) to enable robust, privacy-conscious deployment. The findings aim to guide researchers and practitioners in designing secure, scalable FL systems that improve clinical outcomes while safeguarding patient data and ensuring equitable access across diverse populations.

Abstract

Data privacy has become a major concern in healthcare due to the increasing digitization of medical records and data-driven medical research. Protecting sensitive patient information from breaches and unauthorized access is critical, as such incidents can have severe legal and ethical complications. Federated Learning (FL) addresses this concern by enabling multiple healthcare institutions to collaboratively learn from decentralized data without sharing it. FL's scope in healthcare covers areas such as disease prediction, treatment customization, and clinical trial research. However, implementing FL poses challenges, including model convergence in non-IID (independent and identically distributed) data environments, communication overhead, and managing multi-institutional collaborations. A systematic review of FL in healthcare is necessary to evaluate how effectively FL can provide privacy while maintaining the integrity and usability of medical data analysis. In this study, we analyze existing literature on FL applications in healthcare. We explore the current state of model security practices, identify prevalent challenges, and discuss practical applications and their implications. Additionally, the review highlights promising future research directions to refine FL implementations, enhance data security protocols, and expand FL's use to broader healthcare applications, which will benefit future researchers and practitioners.
Paper Structure (43 sections, 1 equation, 10 figures, 7 tables)

This paper contains 43 sections, 1 equation, 10 figures, 7 tables.

Table of Contents

  1. Introduction
  2. Methodology
  3. Definitions, Theory, and Algorithms
  4. Types of Federated Learning
  5. Horizontal Federated Learning
  6. Vertical Federated Learning
  7. Federated Transfer Learning
  8. Federated Domain Adaption
  9. Multitask Federated Learning
  10. Federated Meta Learning
  11. Model Misconduct
  12. Data leakage
  13. Model poisoning
  14. Model Stealing
  15. Biased Models
  16. ...and 28 more sections

Figures (10)

  • Figure 1: Privacy-focused FL for medical data: This distributed framework allows for the sharing of model weights among different clients. It assesses generalization on decentralized data by combining traditional deep learning with innovative decentralized techniques. This approach is used to improve predictive accuracy and provide detailed insights into the Monkeypox outbreak ahsan2024enhancing.
  • Figure 2: Study selection process depicted using the PRISMA flow diagram, including identification, screening, and inclusion steps.
  • Figure 3: This survey examines the most recent advancements in FL within the field of computer science and engineering, particularly emphasizing the healthcare sector. It presents statistics on (a) the number of papers published in the past five years on FL and (b) the distribution percentage of these papers across various domains. The plots illustrate a steady increase in recent literature.
  • Figure 4: Horizontal FL in the healthcare sector yang2019federated.
  • Figure 5: Vertical FL in the healthcare sector wang2020fed.
  • ...and 5 more figures