Development and Validation of Novel Biomarkers for Carotid Atherosclerosis Quantification and Monitoring using a Carotid Atlas

Description

Carotid atherosclerosis is a major cause of thrombosis and subsequent emboli, which may block one of the cerebral arteries, causing ischemic stroke. Most strokes associated with carotid atherosclerosis can be prevented by lifestyle/dietary changes and medical/surgical treatments. Effects of different treatment strategies are required to be validated in clinical trials. Traditional clinical trials use incidence of vascular events, stenosis and intima-media thickness to evaluate the severity of carotid disease. These trials must be very large in order to demonstrate the treatment effects statistically. Therefore, there is a critical need for sensitive biomarkers for serial monitoring of atherosclerosis. A sensitive, non-invasive 3D imaging technique allowing plaque visualization and quantification will play a vital role in the development of such biomarkers, which may ultimately be used clinically to tailor individual treatment decisions through improved monitoring of treatment response.With the ability to generate 3D carotid images, 3D ultrasound has allowed the development of biomarkers such as total plaque volume and vessel wall volume. However, plaque dimension alone cannot identify the local change in plaque. To address this issue, we developed a method to construct a 3D map to display local vessel-wall-plus- plaque thickness change (VWT-Change) on a point-by-point basis for each artery. Unfortunately, the map depends on the 3D geometry of the carotid wall, which is highly subject-dependent, precluding quantitative comparisons of VWT-Change distribution between subject groups.This proposal describes a technique for mapping VWT-Change computed for different subjects on a standardized 2D carotid atlas, with which the VWT-Change distribution of subjects belonging to different treatment groups can be directly compared. We propose to develop algorithms that utilize the information available on the atlas to: (1) construct two novel biomarkers based on feature selection analysis. A feature is a single VWT-Change measurement at a point on the atlas. The first biomarker is the average VWT-Change over all selected feature points, and the second is based on a discriminant analysis technique between treatment groups. Since regions with selected feature points exhibit greater plaque response difference between treatment groups, we hypothesize that these biomarkers are more sensitive than existing biomarkers; and (2) identify subjects who do not respond to a treatment sufficiently and subjects who experience rapid plaque progression. We hypothesize that these subjects can be identified even in a small pilot clinical study because of the increased sensitivity provided by the carotid atlas. We will validate these hypotheses using images from several clinical studies.