An advanced data fabric architecture leveraging homomorphic encryption and federated learning

Abstract

In this study, we present a novel approach for securely analyzing medical images using federated learning and partially homomorphic encryption within a distributed data fabric architecture. Our approach allows multiple parties to collaboratively train a machine learning model without exchanging raw data, while still maintaining compliance with laws and regulations such as HIPAA and GDPR. We demonstrate the effectiveness of our approach using pituitary tumor classification as a case study, achieving an overall accuracy of 83.31%. However, the primary focus of our work is on the development and evaluation of federated learning and partially homomorphic encryption as tools for secure medical image analysis. Our results show the potential for these techniques to be applied in other privacy-sensitive domains and contribute to the growing body of research on secure and privacy-preserving machine learning.