An MPEG-7 image retrieval system of atherosclerotic carotid plaque images

This paper proposes an MPEG-7-based ultrasound image retrieval system which can assist in studying and analyzing images from atherosclerotic carotid plaques. The proposed system is able to retrieve and sort ultrasound images of the atherosclerotic carotid plaques using image texture analysis features and the KNN classification algorithm. The prototype system was evaluated on a database of 274 images of symptomatic and asymptomatic plaques and the results show the efficiency and efficacy of the proposed approach. The best percentage success rate of correct retrievals was obtained for the NGTDM, followed by the SGLDM (range) feature sets with the corresponding scorings of 70%, and 68%, respectively. The proposed system can be very useful to physicians who, in order to select the best treatment strategy and therapy for a patient, need to find and analyze related cases.