Ethan J. Halpern MD, Presenter: Nothing to Disclose

Eric Laurence Gingold PhD, Abstract Co-Author: Nothing to Disclose

Hugh White MD, Abstract Co-Author: Nothing to Disclose

Katrina M. Read MS, Abstract Co-Author: Employee, Koninklijke Philips Electronics NV

Iterative Model Reconstruction (IMR), as implemented in the second generation Philips software for multislice CT, is a knowledge-based reconstruction with marked reduction in image noise. The purpose of this study was to evaluate image quality with the application of IMR to coronary CT angiography (cCTA).

We evaluated 20 consecutive cCTA cases acquired with a 256-slice iCT scanner, following reconstruction with traditional filtered back projection (FBP), first generation iterative reconstruction (iDose) and second generation IMR (Philips Medical Systems; Cleveland, OH). Each case was reconstructed in a diastolic phase (75% of the R-R interval) and evaluated by two independent reviewers. The mean and standard deviation (sd) of pixel values were computed in a standardized region of interest in the left ventricle and left main coronary artery. Subjective rating scores were obtained from each reviewer (1-5 scale for poor-excellent) for definition of 1) contours of small coronary arteries (<3mm), 2) small coronary calcifications, 3) irregularity/contour of non-calcified plaque, and 4) overall diagnostic confidence for presence/absence of stenosis. The three reconstruction techniques were compared with ANOVA, followed by individual paired comparisons.

There was no significant difference in mean pixel intensity among FBP, iDose and IMR (p>0.8). However, image noise within a contrast-enhanced region of interest in the left ventricle was reduced by a factor of 1.5 from FBP to iDose (sd=123 vs 80,p<0.001) and by a factor of 4.7 from iDose to IMR (sd=80 vs 17,p<0.001). Image quality scores were significantly better with IMR as compared with iDose and FBP (scores for IMR:iDose:FBP in each category: arterial contour: 4.8 > 3.2 > 2.1 (p<0.001); coronary calcification: 4.9 > 3.3 > 2.2 (p<0.001); non-calcified plaque contour: 4.4 > 2.8 > 1.7 (p<0.004); overall diagnostic confidence: 4.9 > 3.8 > 2.6 (p<0.001).

Second generation IMR reduces intravascular noise on cCTA by 86% compared with FBP, allowing better definition of smaller coronary arteries, better discrimination of coronary calcification, better definition of non-calcified plaque, and improved overall diagnostic confidence in the presence or absence of coronary stenosis.

IMR provides images with decreased noise and improved definition of fine anatomical details in the coronary circulation without requiring additional radiation exposure.

Halpern, E, Gingold, E, White, H, Read, K, Clinical Evaluation of Coronary Artery Image Quality with Second Generation Iterative Model Reconstruction .  Radiological Society of North America 2013 Scientific Assembly and Annual Meeting, December 1 - December 6, 2013 ,Chicago IL. http://archive.rsna.org/2013/13016501.html