Segmentation of CT images for efficient fusion with MR images

The fusion of computed tomography (CT) image and magnetic resonance (MR) image is a crucial step in many medical imaging applications. However, fused image quality depends on the fusion methods used to extract relevant information from each modality. This paper presents an image fusion method based on segmentation, and presents the significance of selecting an efficient segmentation method in achieving superior image fusion results. A novel approach is proposed that utilizes Otsu’s Method to segment the skull region from the CT image. Soft tissue region is then extracted from the MR image using the extracted skull region. The skull region and soft tissue region extracted are merged to obtain a fused image. The proposed method is evaluated on a dataset of ten registered pairs of CT and MR images. Superiority of Otsu’s segmentation method over ACM and FCM segmentation methods for the fusion of CT image and MR image is demonstrated using Qualitative and quantitative analysis methods.