Performance Analysis of Contourlet Features with SVM Classifier for the Characterization of Atheromatous Plaque in Intravascular Ultrasound Images

摘要

Medical Image Processing has full-fledged in recent years and demands high accuracy since it deals with human creature. Artificial intelligent is one of the techniques used in this field which aims to reduce human error as much as possible. Hence, in this work, the local characterization of atheromatous plaque is proposed using the feature vector which includes the texture features extracted from the sub bands of third level contourlet transform. The extracted feature vectors are inputted to the SVM Classifier. The classifier differentiates each pixel in the IVUS image as Fibrotic, Lipidic and Calcified plaque tissues. The pixel based classification performance is assessed in terms of sensitivity, specificity and accuracy. The time taken to obtain the average accuracy of 95.92% is about 2 seconds under testing condition