Wisnumurti Kristanto MSC, Abstract Co-Author: Nothing to Disclose

Peter M.A. van Ooijen PhD, Presenter: Nothing to Disclose

Jaap Michiel Groen, Abstract Co-Author: Nothing to Disclose

Rozemarijn Vliegenthart MD, PhD, Abstract Co-Author: Nothing to Disclose

Matthys Oudkerk MD, PhD, Abstract Co-Author: Nothing to Disclose

Previous studies reported the ability of computed tomography (CT) to characterize plaque composition based on their Hounsfield Unit (HU) values. However, it was also reported that HU values of non-calcified plaques were heavily influenced by the lumen contrast-enhancement. This study aimed to evaluate this influence by using software simulation of coronary CT angiography.

Several realistic virtual plaque-burdened coronary phantoms were designed at 5 µm resolution. The phantoms were virtually scanned using the 64-row multi-detector-CT (64-MDCT) clinical scan settings. Scanning and reconstruction were simulated at 64-MDCT and MicroCT resolutions, 0.4 mm and 48 µm, respectively. Influence of lumen HU-values (50 HU and 200-400 HU with 50 HU intervals), stenosis-grades (50% and 75%), and plaque-compositions (fibrous plaque with varied lipid-content: 0-100% with 25% intervals) on plaque HU-values were analyzed. Lumen contrast-enhancement influences were extracted by performing pixel-by-pixel comparisons of contrast-enhanced (200-400 HU) and non-contrast-enhanced (50 HU) images.

Lumen contrast-enhancement influence on plaque HU-value on all dataset followed a specific exponential pattern: y = Ae-λx + C, with x being the distance from the lumen-wall border. The patterns were similar, regardless of the plaque composition, stenosis grade, and the modality. The A determines the influence magnitude. The λ’s, which reflect the range of the influence, were approximately the same for all dataset. The influence was the largest close to the lumen border and exponentially decreasing until 2-pixel radius. Outside this 2-pixel radius, plaque HU-values deviated maximally 5HU. Albeit the variance introduced by the noise, the general patterns of lumen influence remain the same.

To avoid lumen contrast-enhancement influence, plaques have to be measured outside 2-pixel radius from the lumen border. A lumen contrast-enhancement influence correction may be developed based on the discovered patterns.

The patterns of lumen contrast-enhancement influence on non-calcified plaque CT visualization may promote further effort to improve the accuracy of CT-based plaque characterization.

Kristanto, W, van Ooijen, P, Groen, J, Vliegenthart, R, Oudkerk, M, Lumen Contrast Enhancement Influence on Non-calcified Atherosclerotic Plaque CT Visualization.  Radiological Society of North America 2011 Scientific Assembly and Annual Meeting, November 26 - December 2, 2011 ,Chicago IL. http://archive.rsna.org/2011/11012842.html