Abstract: In order to overcome the drawback of the traditional fusion methods that the medical feature information cannot be well represented,a fusion method based on the phase and the magnitude information was proposed in the complex shearlet transform domain.Firstly,the images to be fused were decomposed into lowpass subbands and highpass subbands.Then,a statistical model by employing two-state Von-Mises distribution was constructed to describe the bimodal character of the phase distribution in each highpass subbands.The typical structural similarity of image(SSIM)was extended into the complex shearlet transform domain.Based on the new form of the SSIM and the statistical model,a global-to-local fusion rule was proposed.Finally,the fusion results were obtained by the inversion of the complex shearlet transform.Experimental results demonstrate that better fusion results can be obtained by the proposed method by the comparison of visual sense and typical quantitative measurements.