Neuro-Fuzzy based Multimodal Medical Image Fusion

The role of fusion of images in determining the quality of medical images for diagnosis and treatment planning is important. Present work deals with fusion in wavelet domain with approximation components applied with Principal Component Analysis(PCA) fusion rule and detailed components with Neuro-Fuzzy rule. The pca helps to keep structural information in each position. Neuro-Fuzzy computing allows better intelligent decision-making in the selection of detailed components. The proposed algorithm provides better visualization and tissue structure of the organ. The proposed algorithm can work for both color as well as grayscale images. The proposed fusion algorithm tested with four pairs of data sets which are collected from defense journal. Well-known quantitative measures and subjective assessment by a team of expert radiologists validates the study of the proposed algorithm.