An Improved Medical Image Fusion Algorithm

Abstract

Medical diagnosis becomes widely popular with the progress of technology and social life. High quality medical image fusion technology could provide more effective medical diagnosis image. This paper represents an improved medical image fusion algorithm with three parts. First is image median filtering. Then using modality independent neighborhood descriptor to realize the image registration. Last is the medical image fusion algorithm including the wavelet transform, weighted average and IHS. This paper uses CT, MRI and SPECT images and puts entropy, PSNR, RMSE and average gradient in tests. The results show that the fusion medical image has rich medical information, and it is useful for different medical image fusion scene.