Medical Brain Image Fusion Via Convolution Dictionary Learning

Abstract

Multimodal medical brain-image fusion technology provides effective support for medical diagnosis. This study draws on the global and local advantages of using a convolution dictionary, and proposes an image-fusion technology based on convolution dictionary learning for use on medical images of the brain. A fast Fourier transform is applied to the source image, and the image is decomposed into low-frequency and high-frequency sub-bands; then, suitable fusion rules are used for the low-frequency and high-frequency sub-bands respectively; finally, the inverse fast Fourier transform is used to obtain the fused image. Experimental results show that the proposed fusion technology is superior to the comparison algorithm in objective performance indicators. Moreover, subjective evaluation shows that the texture of the image obtained by the proposed fusion technology is more detailed and more informative.