Tumor detection in manually selected regions of MRI images

Web proceedings papers

Authors

Petre Lameski , Darko Kulakov , Eftim Zdravevski and Andrea Kulakov

Abstract

In this paper we present the results of tumor detection in MRI images using image processing techniques. The data is obtained from MRI scans of patients. The scans are manually divided based on the region of the brain that is scanned. We use several image descriptors and SVM classifier to create the models for positive regions and negative regions (regions that contain tumor and regions that do not contain tumors). A separate SVM is for each descriptor type with and without the region information is trained and the performance of each classifier is analyzed.

Keywords

MRI, Image Processing, Tumor Detection, SVM