Novel extraction and tumour detection method using histogram study and SVM classification

Abstract

In this article we present a new method for detecting and segmenting brain tumour regions weighted brain MRI in T1 (with contrast). This method consists of three main stages: (i) extracting the region of interest (brain) using our EMBE method; (ii) study and histogram analysis of the MRI image to create learning and initialise the classification algorithm will be applied later to retrieve and locate the tumour; (iii) tumour detection and classification using SVM into two classes: tumour class and no-tumour class. Our method will be completed by a characterisation of the tumour area by determining its geometric properties. This work will facilitate later the immense task radiologists to the significant number of MRI images have to deal with daily, and may also be a way for future researchers in order to develop other new methods and develop this research so interesting.