Brain Tumor Segmentation in MRI Images Based on Image Registration and Improved Fuzzy C-Means (IFCM) Method

Registration and Segmentation are important aspects of medical image processing. This paper proposes an efficient Image registration and Improved Fuzzy c-means (IFCM) segmentation technique, to segment the tumor in an MRI medical image. The Affine transform and correlation method have been applied for the image registration, leading to sub entire pixel accuracy for the entire data set. Then, four clustering set models are generated for each registered image in the IFCM based segmentation process. One of the clusters set model is applied to a morphological process, to get the eroded image for extracting the tumor part accurately. The performance of the proposed method is validated both quantitatively and qualitatively, using performance metrics such as standard deviation and entropy.