Wavelet based watermarking approach of hiding patient information in medical image for medical image authentication

Abstract

Watermarking scheme is required to secure data to protect digital contents from unauthorized modifications. Image watermarking scheme can effectively be used in medical image processing to authenticate or investigate the integrity on medical images. In this paper, discrete wavelet transform (DWT) domain and chaotic system based medical image watermarking scheme has been proposed for hiding patient information in medical image to authenticate, in other words, to trace the origin of the image. The patient information has been embedded into the medical images as watermark, without affecting the image quality. DWT has been applied to the host medical image to decompose the image into four sub-bands of low and high frequency components. Chaotic system shows complex behavior and improves the security level of watermarking scheme by providing secret keys. In this paper, logistic map has been used through which chaotic watermark has been obtained, which has been embedded into the low frequency sub-band of medical image. The low frequency sub-band has been divided into 3×3 non overlapping blocks. The chaotic watermark has been embedded into each block by observing neighbor pixels conditions and modifying one of neighbor pixels. Experimental results of the proposed method have been compared with other existing medical image watermarking scheme. From the experimental results, it can be seen that the proposed watermarking scheme outperforms the existing methods. The proposed method achieves high values of peak signal to noise ratio (PSNR) of watermarked image and high values of normalized correlation (NC) of the extracted watermark.