Efficient MRI Brain Image Segmentation and Classification for 3D-Applications

Abstract

New developments in computing power and add-on fabrication makes possible the efficient simulation, development, and replication of proper strategy for patients and prosthetic for neurosurgery applications. Two favorable applications are in finite element modeling for brain injury simulation and recognition and applying stable manufacturing for brain referents. Although these implementations are very impressive, the complications were im-mobile at the segmentation of imaging data for the usage in finite element modeling and printing 3D-techniques. Here, we implemented an innovative algorithm for MRI brain image segmentation that merges analytical segmentation techniques with partial differential equation-based methods using the Adaptive mean shift-modified fuzzy c means (AMS-MFCM) algorithm. We propose a computerized splitting of MRI brain images into several tissue classes using a support vector machine classifier. In the proposed system, the AMS-FCM algorithm enactment for segmentation, and an SVM classifier used for image classification if the given image is normal or abnormal. The Discrete Wavelet Transform is pertained to estimate the factors in the designed methodology.