A novel fusion technique for CT and MRI medical image based on NSST

Abstract

In the field of CT and MRI medical image fusion, For the calculation complexity problem of image fusion based on non-subsampled contourlet transform(NSCT), a novel CT and MRI medical image fusion algorithm is proposed based on non-subsampled shearlet transform (NSST) and compressive sensing(CS) theory. Firstly, NSST is employed to decompose source images respectively, getting one low-pass sub-image and some band-pass directional sub-band images, which have the same size as source images. Secondly, using the improved weighted fusion rule to fuse the low-pass sub-band coefficients, to improve the problem that the contour of the fused image is fuzzy based on the traditional rule, meanwhile, for band-pass directional sub-band coefficients which are featured with high calculation complexity, the fusion rule based on CS is employed, namely, to use compressive sampling technology to sample band-pass directional sub-band coefficients, making a few of observations participate in the calculation of the fusion, to improve the execution speed of codes. Finally, utilizing the inverse NSST to obtain the final fusion image. Experimental results show that the proposed algorithm not only enriches details of the fusion image, but also reduces the calculation complexity, in addition, the proposed improves the problem that the traditional multi-scale decomposition methods for image fusion bring out “Gibbs” effect.