FUSION BASED ENHANCEMENT OF MEDICAL IMAGES

Abstract

CT and MRI imaging modality of medical image, when individually used, cannot provide comprehensive and accurate information, a system to combine (fuse) the two images which give both functional and anatomical information in a single image without any additional artifacts is needed. Image fusion (IF) integrates complementary, as well as redundant information from multiple images to create a fused image, providing more complete and accurate description. Fused image helps radiologist in locating tissue in the brain as well as in deciding the stage of the tumor i.e. if it is malignant or benign. In this report, Image fusion of two medical images viz. CT and MR images is proposed using a novel image fusion method based on segmentation and is compared with the global energy merging scheme which is a region-based analysis approach based on Discrete wavelet transform (DWT). At first, Bone structure from CT image is segregated from the CT image using image segmentation technique and then the segmented CT image and the MR image is fused together to get the fused image. The composite images can preserve more relevant information from the two original images. Experiments showed that the proposed segmentation based image fusion method can have better performance than other DWT based methods and other existing medical image fusion methods.