Critical Challenges and Research Gaps in Blockchain-Based Federated Learning for Healthcare: A Comprehensive Review

The incorporation of federated learning (FL) and blockchain has appeared as a transformative approach for privacy-preserving, decentralised healthcare data sharing. This literature review examines recent advancements in FL-blockchain frameworks applied to healthcare, focusing on critical challenges including data privacy, security, collaboration barriers, centralisation issues and scalability concerns. The findings reveal that blockchain enhances data integrity, access control and trust, while FL enables collaborative model training without sharing raw patient data. Despite these advantages, significant challenges persist such as vulnerabilities to adversarial attacks, regulatory compliance gaps with General Data Protection Regulation (GDPR) and Health Insurance Portability and Accountability Act (HIPAA), handling heterogeneous, non-independent, and identically distributed (non-IID) medical datasets that degrade model performance. The review highlights emerging solutions including secure multi-party computation, smart contract-based aggregation, and incentive mechanisms, while emphasizing the need for future research to develop regulatory-aligned, scalable, and anomaly-resilient FL-blockchain architectures specific to healthcare. Addressing these challenges is essential to establish a trustworthy, efficient, and legally compliant AI-driven healthcare systems.