Diagnosis of astrocytoma and globalastom using machine vision

摘要

Brain tumors are one of the two leading causes of death in the world. A timely and accurate diagnosis of tumor type can help to treatment and recovery greatly. in this paper, we used 30 feature extraction of the MRI images by TCIA center to detect two type ASTROCITOMA and GLIOBLASTOMA tumors. Classifier algorithms as Linear SVM, non-linear SVM and LDA used for division between tumors. Result and study, show than Linear SVM and nonlinear SVM with Quadatic kernel are better than other methods to detect type tumors.