Estudo da técnica de VBM em imagens de ressonância magnética de cérebros com lesões

Abstract

This project deals with the study and application of the technique VBM (Voxel Based Morphometry) on magnetic resonance images (MRI) of brain with lesions, which involves advanced techniques of Digital Image Processing (DIP). The technique of VBM is based on comparison between voxel to voxel images of two individuals or groups of individuals. This comparison is related to the local concentration of gray levels. It is a process that involves spatial normalization followed by segmentation of images normalized and smoothed (smoothing) of the substances targeted. A statistical map is made using the SPM (Statistical Parametric Mapping), according to comparisons made between the targeted substances, getting through this comparison, for example, the composition of substances brain (gray, white, or CSF), which can be an aid in the diagnosis of various neurological diseases. However, brain lesions invalidate this technique, because the presence of a lesion in region generates a big problem during the steps of normalization and segmentation, since these processes take into account global information from the brain. Thus the objective is to implement a modified version of VBM for the SPM, taking into account images of brain lesions, using advanced digital image processing to achieve satisfactory results in comparison of those brains.