Intra-Vascular Ultrasound Image Based Decision Support System For Coronary Plague Classification

Abstract

Heart attacks continue to be one of the most common reasons for increasing mortality now-a-days. Intra-Vascular UltraSound (IVUS) imaging has recently been used in the development of a variety of classification methods for the plaques that are seen in the coronary arteries. Plaque analysis is an important tool for early detection of coronary artery disease, which may have devastating consequences. A powerful decision support system (PDSS) for the plaque categorization using IVUS image is presented in this paper. Multi Directional Transform (MDT) is employed for extracting features and Maximum Likelihood Classifier (MLC) is employed for the classification. In order to remove the speckle noise from the given IVUS image, a preliminary processing step is performed on it. After the preprocessing step has been finished, the proposed system will extract features using MDT, and it will then classify the IVUS images as either normal or abnormal. Based on the results of the evaluation of the proposed system, it seems that the classification of IVUS images may be accomplished exclusively by the extracted texture features, and that the accuracy of the classifier is dependent on these derived features. It has been shown that the system produces a satisfactory outcome, with a sensitivity of 98%.