SVM CLASSIFIER FOR THE IDENTIFICATION OF ABDOMINAL AORTA ANEURYSM

The local enlargement of abdominal aorta creates severe threat to older people and smokers once it burst. As it grows without any symptoms, specialists can diagnose the existence of such aneurysm by utilizing abdominal ultrasound scanning. This process relies on the position and sizing hence accuracy of the detection is significantly important. To reduce the devastating effects of abdominal aortic aneurysm (AAA) during the medical treatment, several researchers have proposed various novel ideas and algorithms to get the infallible result from the successful treatment of AAA. In this paper, the processing and filtering of input AAA image is done by median channel followed by Watershed Transform based segmentation is done to segregate the aneurysm portion from the remaining abdomen area. The features selected by GRCM are classified by Support vector machine. The accuracy and sensitivity are improved with the proposed approach with the maximum accuracy of 91% and sensitivity 95%. When compared with conventional method, the implementation of the proposed method is facile with superior performance.