Automatic Brain Tumor Extraction for MRI-T1 and T2 using Geodesic Distance and Statistical Methods

In this paper, we present a new approach that allows the detection of brain tumors. The approach is based on mathematical methods such as correlation, covariance and geodesic distance. Before proceeding to the segmentation and automatic extraction, the detection of central indices of abnormal tissues is based on the method of correlation and covariance. From these indices, segmentation of brain tumor area using geodesic distance in T1 and T2 magnetic resonance images (MRI-T1 and T2). The ultimate objective is to retrieve the attributes of the tumor observed on the image; these attributes form a characterizing vector, which is used latter in the extraction and classification steps to get a better diagnosis. The proposed method yielded fruitful results and has shown a better performance in the analysis of biomedical images of modality MRI-T1 and T2. Keywords— Biomedical Images Processing, Detection, Segmentation, Correlation, Covariance, Geodesic Distance.