Abstract: The accelerated growth of Artificial Intelligence (AI) has greatly influenced the healthcare industry, offering significant advancements in smart healthcare systems. However, the lack of standards, legal regulations, and is difficult to meet ethical standards for patient information privacy. The utilization of large quantities of user data for training machine learning models has shown promising results. Nonetheless, two major obstacles persist: the fragmented nature of user data, hindering aggregation without compromising privacy, and the failure of cloud-based models to personalize healthcare. To address these issues, Federated Learning (FL) has emerged as a solution, leveraging privacy-preserving algorithms to overcome data atomization concerns. Furthermore, integrating FL with technologies like blockchain and edge computing can enhance security and computational efficiency.
This paper presents an overview of FL architectures, comparing many kinds of federated learning frameworks and distributed machine learning algorithms. It explores the limitations of current smart healthcare systems and highlights how FL can overcome these challenges. The study investigates different FL architectures and classification models, showcasing their potential application in healthcare.
Furthermore, it analyses the advantages of FL in medical settings, emphasizing privacy preservation and improved data management. The paper also assesses the security risks associated with healthcare applications and proposes ways to mitigate them. The research findings aim to help both academia and industry understand the competitive advantage offered by advanced privacy-preserving federated learning systems in the field of healthcare data.
Keywords: Artificial Intelligence (AI), Federated Learning, Privacy Preservation, Data Management, Security Risks.
Title: A Survey on Federated Learning for Intelligent Healthcare Systems
Author: Mrs. Deepthi S, Dr.R.Chinnayan
International Journal of Engineering Research and Reviews
ISSN 2348-697X (Online)
Vol. 11, Issue 4, October 2023 - December 2023
Page No: 8-17
Research Publish Journals
Website: www.researchpublish.com
Published Date: 25-October-2023
