Fusion of Medical Images Using Mutual Information and Intensity Based Image Registration Schemes

Image fusion is the process of combining two or more images for providing more information. Medical image fusion refers to the fusion of medical images obtained from different modalities. Medical Image Fusion helps in medical diagnosis by way of improving the quality of the images. In diagnosis, images obtained from a single modality like MRI, CT etc, may not be able to provide all the required information. It is needed to combine information obtained from other modalities also to improve the information acquired. For example combination of information from MRI and CT modalities gives more information than the individual modalities separately. The aim is to provide a method for fusing the images from the individual modalities in such a way that the fusion results in an image that gives more information without any loss of the input information and without any redundancy or artifacts. In the fusion of medical images obtained from different modalities they might be in different coordinate systems and they have to be aligned properly for efficient fusion. The aligning of the input images before proceeding with the fusion is called image registration. The intensity based registration and Mutual information based image registration procedures are carried out before decomposing the images. The two imaging modalities CT and MRI are considered for this study. The results on CT and MR images demonstrate the performance of the fusion algorithms in comparison with registration schemes.