Performance evaluation of frequnecy domain techniques for efficient storage of patient information with medical images

Abstract

In an e-Health environment medical image protection and authentication are extremely important as medical image and medical data are distributed over electronic network. Research has shown that medical image watermarking is relevant process for enhancing data security and content authentication. Medical image has most sensitive data hence while adding the external data it is important to preserve the perceptual integrity of an image. This Paper presents comparative analysis of interleaving patient information in the form text and electrocardiograph (ECG) signal in medical images, using Discrete Fourier Transform (DFT) and Discrete Cosine Transform (DCT). To provide the security, text data and ECG signal are encrypted before interleaving in the frequency domain. Predicative coding techniques like adaptive delta modulation and differential pulse code modulation are used for ECG signal encryption and compression. The quantitative assessment of the quality of watermarked images is carried out using peak signal to noise ratio (PSNR) and normalized mean square error (NRMSE). The experimental results on different medical image modalities (MRI, MRA, CT) are tabulated, which shows that the DFT based interleaving algorithm is efficient as it preserves imperceptibility better than DCT.