Federated Learning for Privacy Preservation in Smart Healthcare Systems: A Comprehensive Survey

Recent advances in electronic devices and communication infrastructure have\nrevolutionized the traditional healthcare system into a smart healthcare system\nby using IoMT devices. However, due to the centralized training approach of\nartificial intelligence (AI), the use of mobile and wearable IoMT devices\nraises privacy concerns with respect to the information that has been\ncommunicated between hospitals and end users. The information conveyed by the\nIoMT devices is highly confidential and can be exposed to adversaries. In this\nregard, federated learning (FL), a distributive AI paradigm has opened up new\nopportunities for privacy-preservation in IoMT without accessing the\nconfidential data of the participants. Further, FL provides privacy to end\nusers as only gradients are shared during training. For these specific\nproperties of FL, in this paper we present privacy related issues in IoMT.\nAfterwards, we present the role of FL in IoMT networks for privacy preservation\nand introduce some advanced FL architectures incorporating deep reinforcement\nlearning (DRL), digital twin, and generative adversarial networks (GANs) for\ndetecting privacy threats. Subsequently, we present some practical\nopportunities of FL in smart healthcare systems. At the end, we conclude this\nsurvey by providing open research challenges for FL that can be used in future\nsmart healthcare systems\n