A Novel Reversible Multiple Medical Image Watermarking for Health Information System.

Abstract

The exponential augmentation of information and communication technologies has appreciably shifted panorama of Health Information System (HIS), drawing keen interests of the researchers to address the challenges in protecting the medical data. Digital Watermarking therefore gained wide popularity as an entrusted measure to ensure the security. Literature reports wide assortment of imaging approaches for digital watermarking; however its perspectives in the HIS have received meager attention, especially in the reversible watermarking context. This paper presents wavelet based novel reversible watermarking approach which imperceptibly embeds multiple watermarks like patient diagnosis detail, Electrocardiogram (ECG) signal, doctor's digital signature and diagnostic code for image retrieval. The proposed wavelet based approach exploits the quantization function to embed multiple watermarks and uses a novel tracking key to recover the original medical image. The algorithm has been analyzed on various medical imaging modalities (MRI, CT, MRA, US images of size 512 X 512), considering various image processing attacks. Experimental results signify highest hiding capacity of 91,136 bits with PSNR of 44.52dB. The proposed algorithm adheres to strict necessities concerning the tolerable modifications in the medical images. The technique addresses the issues related with HIS namely data integrity, authentication, confidentiality protection, efficient data management, retrieval of image and additionally fulfill tamper detection on the entire image for ensuring the integrity of the medical image.