Share Email Print
cover

Proceedings Paper

Computerized flow and vessel wall analyses of coronary arteries for detection of non-calcified plaques in coronary CT angiography
Author(s): Jun Wei; Chuan Zhou; Heang-Ping Chan; Aamer Chughtai; Prachi Agarwal; Lubomir Hadjiiski; Ella Kazerooni M.D.
Format Member Price Non-Member Price
PDF $14.40 $18.00
cover GOOD NEWS! Your organization subscribes to the SPIE Digital Library. You may be able to download this paper for free. Check Access

Paper Abstract

The buildup of non-calcified plaques (NCP) that are vulnerable to rupture in coronary arteries is a risk for myocardial infarction. We are developing a computer-aided detection (CADe) system to assist radiologists in detecting NCPs in cCTA. A major challenge of NCP detection is the large number of false positives (FPs) caused by the small sized coronary arteries, image noise and artifacts. In this study, our purpose is to design new image features to reduce FPs. A data set of 98 cCTA scans was retrospectively collected from patient files. We first used vessel wall analysis, in which topological features were extracted from vessel wall and fused with a support-vector machine, to identify the NCP candidates from the segmented coronary tree. Computerized flow dynamic (CFD) features that characterize the change in blood flow due to the presence of plaques and a vascular cross-sectional (VCS) feature that quantifies the presence of low attenuation region at the vessel wall were designed for FP reduction. Using a leave-one-out resampling method, a support vector machine classifier was trained to merge the features into a NCP likelihood score using the vessel wall features alone or in combination with the new CDF and VCS features. The performance of the new features in classification of true NCPs and FPs was evaluated by the area under the receiver operating characteristic (ROC) curve (AUC). Without the new CFD and VCS features, the test AUC was 0.84±0.01. The AUC was improved to 0.88±0.01 with the addition of the new features. The improvement was statistically significant (p < 0.001). The study indicated that the new flow dynamic and vascular cross-sectional features were useful for differentiation of NCPs from FPs in cCTA.

Paper Details

Date Published: 24 March 2016
PDF: 6 pages
Proc. SPIE 9785, Medical Imaging 2016: Computer-Aided Diagnosis, 97853U (24 March 2016); doi: 10.1117/12.2216815
Show Author Affiliations
Jun Wei, Univ. of Michigan Health System (United States)
Chuan Zhou, Univ. of Michigan Health System (United States)
Heang-Ping Chan, Univ. of Michigan Health System (United States)
Aamer Chughtai, Univ. of Michigan Health System (United States)
Prachi Agarwal, Univ. of Michigan Health System (United States)
Lubomir Hadjiiski, Univ. of Michigan Health System (United States)
Ella Kazerooni M.D., Univ. of Michigan Health System (United States)


Published in SPIE Proceedings Vol. 9785:
Medical Imaging 2016: Computer-Aided Diagnosis
Georgia D. Tourassi; Samuel G. Armato III, Editor(s)

© SPIE. Terms of Use
Back to Top