Multimodal medical image fusion in nonsubsampled contourlet transform domain

Abstract

This paper proposed a novel medical image fusion algorithm based on nonsubsampled contourlet transform (NSCT). Firstly, CT image and MRI image are decomposed into high and low frequency subband by NSCT. For selection of fusion rules, the coefficients of low frequency subband are fused by local entropy, and fusion criterion for the coefficients of high frequency subband is regional energy. The performance of the proposed framework is evaluated by various quantitative measures. Visual and quantitative analysis and comparisons of the proposed algorithm with the existing fusion methods show the effectiveness of the proposed scheme in the application of medical image fusion.