Medical Image Analysis Using Federated Learning Frameworks: Technical Review

Learning from various types of Medical Imaging data have been a area of research since it involve variation in data distribution bias factor, privacy and legal issues. Hence ML algorithms can not be generalized to take care diversity in data-sets taken from various organizations hospitals locations or devices. The federated learning (FL) paradigm helps to overcome such challenges and allowing the algorithms to learn on various medical imaging data sets from various locations through collaborations at central server. In this paper we are presenting a technical review on federated learning architectures for medical image analysis. The manuscript illustrate how federated learning architectures applied for medical image analysis in effective way.