Atherosclerosis Disease Diagnosis by Using SVM and ANN

Shaker K. Ali, Hazeem B. Taher and Baidaa M. Alsafy


Abstract:

Atherosclerosis is a specific form of arteriosclerosis in which an artery wall thickens as a result of invasion and accumulation of white blood cells and proliferation of intimal-smooth-muscle cell creating an athermanous plaque. It is also known as arteriosclerotic vascular disease. It is a chronic disease that remains asymptomatic for decades. There are two types of this disease: the total Occlusion of the artery and the Stenosis the artery. In this paper we suggest anew algorithm for Atherosclerosis diagnoses by using SVM and ANN. Our algorithm includes five stages: first is the preprocessing stage to remove noise and increase image contrast. The second stage is the segmentation stage which segments input images into objects by converting the background to black color, while the artery is to white color.After the segmentation stage is the features extraction stage where it extracts eight features: five shape features(area, Perimeter, Eccentricity, Diameter, Perimeter-to-Area Ratio) and three texture features (Coarseness, Contrast and Directionality). The extracted features are passed to the fourth stage which is the decision making. In our system, the decision making stage is built in a hierarchy of two levels. The first level has SVM to classify the input images into normal or abnormal. Then, the Artificial neural network (ANN) is used in the second level to diagnosis the two types of abnormal Atherosclerosis (Occlusion and Stenosis). The experimental results of our algorithm show 98% accuracy for normal and abnormal artery diagnosis using SVM and 90% accuracy for Occlusion and Stenosis Atherosclerosis diagnosis using ANN.

Volume: 10 | Issue: 11

Pages: 411-417

Purchase this Article