Share Email Print

Proceedings Paper

Cardiac CT estimability index: an ideal estimator in the presence of noise and motion
Author(s): Taylor Richards; Alex Ivanov; Paul Segars; Udo Hoffmann; Ehsan Samei
Format Member Price Non-Member Price
PDF $17.00 $21.00

Paper Abstract

Estimating parameters of clinical significance, like coronary stenosis, accurately and precisely from cardiac CT images remains a difficult task as image noise and cardiac motion can degrade image quality and distort underlying anatomic information. The purpose of this study was to develop a computational framework to objectively quantify stenosis estimation task performance of an ideal estimator in cardiac CT. The resulting scalar figure-of-merit, the estimability index (e’), serves as a cardiac CT specific task-based measure of image quality. The developed computational framework consisted of idealized coronary vessel and plaque models, asymmetric motion point spread functions (mPSF), CT image blur (MTF) and noise operators (NPS), and an automated maximum-likelihood estimator (MLE) implemented as a matched template squared-difference operator. Using this framework, e’ values were calculated for 131 clinical case scenarios from the Prospective Multicenter Imaging Study for Evaluation of Chest Pain (PROMISE) trial. The calculated e’ results were then utilized to classify patient cases into two exclusive cohorts, high-quality and low-quality, characterized by clinically meaningful differences in image quality. An e’ based linear classifier categorized the 131 patient datasets with an AUC of 0.96 (6 false-positives and 10 false-negatives), compared to an AUC of 0.89 (4 false-positives and 20 false-negatives) for a linear classifier based on contrast-to-noise ratio (CNR). In summary, a computational framework to objectively quantify stenosis estimation task performance was successfully implemented and was reflective of clinical results in the context of subset of a large clinical trial (PROMISE) with diverse sites, readers, scanners, acquisition protocols, and patient types.

Paper Details

Date Published: 1 March 2019
PDF: 9 pages
Proc. SPIE 10948, Medical Imaging 2019: Physics of Medical Imaging, 109485W (1 March 2019); doi: 10.1117/12.2513035
Show Author Affiliations
Taylor Richards, Carl E. Ravin Advanced Imaging Labs. (United States)
Alex Ivanov, Massachusetts General Hospital (United States)
Paul Segars, Carl E. Ravin Advanced Imaging Labs. (United States)
Udo Hoffmann, Massachusetts General Hospital (United States)
Ehsan Samei, Carl E. Ravin Advanced Imaging Labs. (United States)

Published in SPIE Proceedings Vol. 10948:
Medical Imaging 2019: Physics of Medical Imaging
Taly Gilat Schmidt; Guang-Hong Chen; Hilde Bosmans, Editor(s)

© SPIE. Terms of Use
Back to Top
Sign in to read the full article
Create a free SPIE account to get access to
premium articles and original research
Forgot your username?