In the field of medicine, multimodal image analysis is attaining importance due to the fact that large number of images with clinical data has to be examined to analyze different types of results. Fusion of multimodal images merges required details from a single or multiple images into a solitary image. Fusion provides increased clinical applicability of medical images which aids in the diagnosis of diseases. Segmentation of fused images will help in identifying meaningful objects in an image based on the problem being solved. Multimodal image fusion is carried using curvelet transform and MSVD (Multi-resolution Singular Decomposition) method. The fused images are segmented using various segmentation techniques such as K means clustering, Mean shift segmentation and Normalized cut segmentation. The performance of various segmentation methods are analyzed using different metrics such as entropy, PSNR (Peak Signal to Noise Ratio) and RMSE (Root Mean Square Error).
. et al. (Sun,) studied this question.