An Efficient Framework to Protect Medical Images

Abstract

In this work, lossless compression is considered and an efficient Encryption–then-compression (ETC) scheme is proposed for medical images. A considerably good level of security is achieved by the proposed scheme where image encryption is operated in the prediction error domain. In addition, reasonably good compression of the encrypted medical images is achieved. Comparison is done between Gradient adjusted prediction (GAP) & Median Edge Detector (MED) techniques and GAP is found to perform better with respect to PSNR, bit rate and entropy. The average values of PSNR, bit rate and entropy in decibels for MED predictor are 10.8313, 3.7695 and 3.8624 respectively. Similarly, for GAP predictor, the average values of PSNR, bit rate and entropy in decibels are 11.9025, 4.0279 and 4.1286 respectively.