Collaborative Screening of COVID-19-like Disease from Multi-Institutional Radiographs: A Federated Learning Approach

Abstract

COVID-19-like pandemics are a major threat to the global health system have the potential to cause high mortality across age groups. The advance of the Internet of Medical Things (IoMT) technologies paves the way toward developing reliable solutions to combat these pandemics. Medical images (i.e., X-rays, computed tomography (CT)) provide an efficient tool for disease detection and diagnosis. The cost, time, and efforts for acquiring and annotating, for instance, large CT datasets make it complicated to obtain large numbers of samples from a single institution. However, owing to the necessity to preserve the privacy of patient data, it is challenging to build a centralized dataset from many institutions, especially during a pandemic. Moreover, heterogeneity between institutions presents a barrier to building efficient screening solutions. Thus, this paper presents a fog-based federated generative domain adaption framework (FGDA), where fog nodes aggregate patients’ data necessary to collaboratively train local deep-learning models for disease screening in medical images from different institutions. Local differential privacy is presented to protect the local gradients against attackers during the global model aggregation. In FGDA, the generative domain adaptation (DA) method is introduced to handle data discrepancies. Experimental evaluation on a case study of COVID-19 segmentation demonstrated the efficiency of FGDA over competing learning approaches with statistical significance.

An Extensive Study and Review of Privacy Preservation Models for the Multi-Institutional Data

“…Stringent data sovereignty laws mandate local storage, hindering seamless cross-border data transfer. Collaboration among healthcare entities and policymakers fosters standardized cross-border datasharing protocols [26] [29].…”

“…In recent years, the integration of deep learning models into clinical applications has ushered in a new era of healthcare innovation. These advanced computational tools have demonstrated immense potential in revolutionizing disease diagnosis, treatment planning, and patient care [26]. The ability of deep learning algorithms to analyze complex medical data, such as medical images, genomic sequences, and electronic health records, has opened up a plethora of opportunities for improving clinical outcomes and healthcare delivery [27].…”

“…According to a recent study conducted IoT devices are increasingly becoming pervasive in our everyday life, so it is important to understand the underpinning privacy and security risks associated with them [26]. These risk factors result in attacks by cyber attackers faced by consumers using IoT devices.…”