MRI Images multi-level Thresholding Based on PSO: Real-Time Hardware Implementation
Fayçal Hamdaoui, Anis Sakly, Abdellatif Mtibaa · International journal of imaging and robotics · 2016
MRI images provide useful and valuable amount of information’s about brain anatomy allowing doctors to analyze and diagnose diseases. To successfully succeed the diagnosis of a brain tumor, several types of sequences and images will be taken and their processes require massive time. This diagnosis will detect, locate, measure and evaluate the consequences of brain tumor. MRI brain image reproduces the internal structure of the human head. It is a multimodal structure. It is mainly composed of three regions: the gray matter, white matter, and cerebrospinal fluid. Therefore, tumor detection requires segmenting MRI image into levels equal to region numbers of MRI image. In this paper, we try to provide an automatic hardware tool for the diagnosis of brain structures using the Xilinx System Generator (XSG) which consists of a realtime MRI images multi-level thresholding based on hybridization of PSO and Otsu method. The performances of the proposed hardware architecture are demonstrated and validated using a set of MRI medical images