Generation of SVD Based Non-informative Unique ID for Authentication of MRI Images

Abstract

This paper puts forth a viable method to generate a non-informative unique ID to overcome the security problems of images in tele-medicine applications. Unlike the prevalent methods for image security like steganography, the proposed scheme does not tamper with the medical images and ensures no loss of medical information. The digital unique ID is obtained by strategically combination of local features extracted from MRI image in Singular Value Decomposition (SVD) domain and patient's information. For authentication of medical image, the patient's information can be recovered from generated unique ID without any loss only after access of the particular unique ID. The proposed method has been tested on the publicly accessible dataset of medical MRI images and promising results were observed for solving the security issues of medical MRI images. Therefore the proposed method can be implemented in the field of tele-medicine and can be applied in areas where accurate and loss-less verification is a must as this may have great relevance when human health is concerned.