Federated Learning Based Disease Detection in Healthcare: A Systematic Review
Abstract
Federated learning (FL) is an approach of collaborative learning which have become the emerging trend in healthcare research. FL effectively tackle the critical issues faced by classical machine learning models, such as clinician data privacy issues, the heterogeneity of data across institutions and regulations. This systematic review work explores the Federated Learning as a distributed solution for collaborative model training in multi-healthcare institution scenario which would not expose raw patient data and its privacy. We carried out a systematic review of the literature to identify articles on FL applications in disease detection from 2020 onwards. We analyzed the three fundamental algorithms namely, FedAvg, FedProx and SCAFFOLD in different use-case scenarios with medical data. Furthermore, we analyzed various models to detect the disease implemented via federated approach by assessing the accuracy, clinical domain, sample size and methods. Major findings show that on heterogeneous data and network settings, FedProx and SCAFFOLD outperform FedAvg in terms of accuracy. This review presents several limitations and research gaps such as the data heterogeneity, non-IID (non-Independent and Identically Distributed) distributions of data, computational requirements, model interpretability. Our survey suggests that although FL provides a promising platform for secure and privacy-preserving collaborative diagnostics, further studies are needed to investigate scalability,…
How strong is the evidence?
Evidence Hub reads this as a systematic review, done in people. Its evidence level is 1 of 5 (Highest · pooled evidence). Pooled evidence from several studies sits at the top of the evidence pyramid.
Read the paper
Related studies
- Active Learning during Federated Learning for Object Detection
- Fairness-aware federated learning
- Decentralized Federated Learning
- Building a Personally Identifiable Information Recognizer in a Privacy Preserved Manner Using Automated Annotation and Federated Learning
- Preservation of Patient Level Privacy: Federated Classification and Calibration Models
- Efficient explainable federated learning
Go deeper in Evidence Hub
Evidence Hub adds a study data sheet (design, sample size, evidence level, key numbers), an AI summary and a chat grounded in the paper's text.