An Imperceptible Secure Transfer of Medical Images for Telemedicine Applications

Abstract

Medical image transfer is an important task to transfer the patient’s medical observations to the doctors at remote areas to know patient condition through the Internet/open network for expert opinion. But information transmission through unsecured channel raises confidentiality problems for any image data. This paper addresses the issue of confidentiality for medical images. The proposed technique uses a hybrid combination of Stationary Wavelet Transformation (SWT) and Singular Value Decomposition of the original medical image for the embedding of Patient watermark into it. The robustness is improved by verifying the peak signal to noise ratio to avoid wrong diagnosis of medical conditions. The robust performance of the proposed scheme is evaluated for known common processing attacks such as Salt and pepper, Gaussian, Mean and Median Filters, Histogram Equalization etc. The method is robust for embedded watermark at the acceptable visual quality of the watermarked image.