Structural medical image fusion by means of dual tree complex wavelet

This paper presents a new method for fusion of medical images. It has been shown that 2-D wavelets are not sufficient for decomposition of images due to lack of directionality and shift variance. Complex wavelet transforms stand as a candidate to overcome deficiencies of wavelet transform. This paper exploits dual tree complex wavelet transform to decompose image and fuse the details and approximations based on maxima and averaging fusion rules. Results of visual and quantitive analysis prove efficiency of the method over conventional multiresolution methods in terms of mutual information, entropy and standard deviation.