Abstract
The proliferation of electronic health records (EHRs) and medical Internet of Things (IoT) data presents an unprecedented opportunity to advance healthcare through data-driven analytics, particularly with deep learning models. However, the sensitive nature of health data, coupled with stringent privacy regulations like HIPAA and GDPR, often isolates data in siloed institutions, creating a significant barrier to developing robust, generalized models. Federated Learning (FL) has emerged as a promising decentralized machine learning paradigm that enables model training across multiple data sources without sharing the raw data. This paper explores the application of FL in the healthcare domain, focusing on its role in preserving patient privacy. We provide a comprehensive literature survey of the current state-of-the-art. The core of this work involves a detailed methodology discussing six prominent federated learning models: Federated Averaging (FedAvg), Federated Averaging with Secure Ag