Counterlet Based Medical Image Compression Using Improved EZW

This paper presents a new coding technique based on image compression using contourlet transform used in different modalities of medical imaging. Recent reports on natural image compression have shown superior performance of contourlet transform, a new extension to the wavelet transform in two dimensions using nonseparable and directional filter banks. As far as medical images are concerned the diagnosis part (ROI) is of much important compared to other regions. Therefore those portions are segmented from the whole image using neural network based fuzzy logic technique. Contourlet transform is then applied to ROI portion which performs Laplacian Pyramid (LP) and directional filter banks to the resultant because of directionality and anisotropy. The region of less significance are compressed using Discrete Wavelet Transform and finally modified embedded zerotree wavelet algorithm is applied which uses six symbols instead of four symbols used in Shapiro's EZW to the resultant image which shows better PSNR and high compression ratio and finally Huffman coding is applied to get the compressed image.