Medical image enhancement based on NSCT

A novel method for medical images enhancement based on NSCT and image statistical properties was proposed. Firstly, nonsubsampled contourlet transform provides a multiresolution and multi-direction analysis for the medical images. We use the nonsubsampled Laplace Pyramids to decompose the image into different scales, then use the nonsubsampled directional filter banks to decompose the image into different directions and estimate the noise variance of each subband. After modeling the coefficient of each directional subband and using generalized classification model according to the statistical property of the transform coefficients, they were categorized into strong edges, weak edges and noise by Bayesian classifier. A nonlinear mapping function was designed to enhance and suppress the different coefficients adaptively to obtain a good enhancement result with significant characters according to the modified coefficients of NSCT. Experimental results show that the proposed method is practicable and effective.