An Imperceptible Medical Image Watermarking Framework for Automated Diagnosis of Retinal Pathologies in an eHealth Arrangement

Abstract

The newly driven concept of electronic health (eHealth) has made distance irrelevant in healthcare. eHealth, which is commonly known as an electronic form of healthcare, refers broadly to the use of advanced information and communication technology within healthcare environments. However, its practical implementation poses many sets of challenges. One of the major challenges is authentication and integrity verification of vital medical data. Proper verification of patients with associated medical images or data is crucial to conduct accurate medical diagnostics. This paper presents an imperceptible watermarking-based security framework to address the issues of authentication and integrity verification of medical images for eHealth applications. The electronic patient record (EPR), as a watermark image, is embedded in optical coherence tomography (OCT)/fundus scan (consisting of healthy and diseased scans), using a hybrid watermarking technique based on fast curvelet transform (FCT) and singular value decomposition (SVD). The proposed work showed a high level of robustness, imperceptibility, and security for medical images when compared with the state-of-the-art existing watermarking techniques. In addition, we conducted a comparative analysis of the watermarked OCT/fundus scans with non-watermarked scans, and our results validate that the insertion of the watermark in the retinal images does not affect the automated medical image diagnosis of various retinal pathologies. Moreover, the correct recovery of the EPR from the watermarked scans makes the proposed framework applicable to the authentication of medical images for a computer-based automated diagnostic system in an eHealth arrangement.