Article,

Towards A More Secure Web Based Tele Radiology System: A Steganographic Approach

, , and .
International Journal of Computer Science and Security (IJCSS), 12 (4): 55 - 75 (December 2018)

Abstract

While it is possible to make a patient's medical images available to a practicing radiologist online e.g. through open network systems inter connectivity and email attachments, these methods don't guarantee the security, confidentiality and tamper free reliability required in a medical information system infrastructure. The possibility of securely and covertly transmitting such medical images remotely for clinical interpretation and diagnosis through a secure steganographic technique was the focus of this study. We propose a method that uses an Enhanced Least Significant Bit (ELSB) steganographic insertion method to embed a patient's Medical Image (MI) in the spatial domain of a cover digital image and his/her health records in the frequency domain of the same cover image as a watermark to ensure tamper detection and nonrepudiation. The ELSB method uses the Marsenne Twister (MT) Pseudo Random Number Generator (PRNG) to randomly embed and conceal the patient's data in the cover image. This technique significantly increases the imperceptibility of the hidden information to steganalysis thereby enhancing the security of the embedded patient's data. In measuring the effectiveness of the proposed method, the study adopted the Design Science Research (DSR) methodology, a paradigm for problem solving in computing and Information Systems (IS) that involves design and implementation of artefacts and methods considered novel and the analytical testing of the performance of such artefacts in pursuit of understanding and enhancing an existing method, artefact or practice. The fidelity measures of the stego images from the proposed method were compared with those from the traditional Least Significant Bit (LSB) method in order to establish the imperceptibility of the embedded information. The results demonstrated improvements of between 1 to 2.6 decibels (dB) in the Peak Signal to Noise Ratio (PSNR), and up to 0.4 MSE ratios for the proposed method.

Tags

Users

  • @cscjournals

Comments and Reviews