Brain tumour diagnostic segmentation based on optimal texture features and support vector machine classifier
by Ahmed Kharrat; Mohamed BenMessaoud; Mohamed Abid
International Journal of Signal and Imaging Systems Engineering (IJSISE), Vol. 7, No. 2, 2014

Abstract: This paper presents a new general automatic method for segmenting brain tumours in Magnetic Resonance (MR) images. Our approach addresses all types of brain tumours. The proposed method involves, therefore, image pre-processing, feature extraction via the wavelet transform-spatial gray level dependence matrix (WT-SGLDM), dimensionality reduction using the Genetic Algorithm (GA) and classification of the reduced features using a support vector machine (SVM). These optimal features are employed for the segmentation of brain tumour. The resulting method is aimed at early tumour diagnostics support by distinguishing between brain tissue, benign tumour tissue and malignant tumour tissue. The segmentation results in different types of brain tissues that are evaluated by comparison with manual segmentation, as well as with other existing techniques. The quantitative evaluation shows that our approach outperforms manual segmentation with match percent (MP) measures equal to 97.08% and 98.89% for malignant and the benign tumours, respectively. The qualitative evaluation displays that our attitude overtakes the FCM algorithm with an accuracy rate of 99.69% for benign tumours and 99.36% for malignant tumours.

Online publication date: Mon, 27-Oct-2014

The full text of this article is only available to individual subscribers or to users at subscribing institutions.

 
Existing subscribers:
Go to Inderscience Online Journals to access the Full Text of this article.

Pay per view:
If you are not a subscriber and you just want to read the full contents of this article, buy online access here.

Complimentary Subscribers, Editors or Members of the Editorial Board of the International Journal of Signal and Imaging Systems Engineering (IJSISE):
Login with your Inderscience username and password:

Username:        Password:          Forgotten your password?

Want to subscribe?
A subscription gives you complete access to all articles in the current issue, as well as to all articles in the previous three years (where applicable). See our Orders page to subscribe.

If you still need assistance, please email subs@inderscience.com