Petra Kozma, Presenter: Nothing to Disclose

Marc Dewey MD, Abstract Co-Author: Research Grant, General Electric Company Research Grant, Bracco Group Research Grant, Guerbet SA Research Grant, Toshiba Corporation Speakers Bureau, Toshiba Corporation Speakers Bureau, Guerbet SA Speakers Bureau, Bayer AG Consultant, Guerbet SA Author, Springer Science+Business Media Deutschland GmbH Editor, Springer Science+Business Media Deutschland GmbH Institutional research agreement, Siemens AG Institutional research agreement, Koninklijke Philips NV Institutional research agreement, Toshiba Corporation

Accuracy of computed tomography angiography (CTA) for coronary stenosis quantification has been reported to be poor when compared with the reference gold standard quantitative coronary angiography (QCA).

A moving heart phantom with non-calcified coronary stenoses of different size (25, 50, and 75%) as well as different shapes (D-shaped, concentric, and eccentric) were imaged using 320-row CTA. Image acquisition parameters were: 80, 10, and 120 kVp with tube currents of 100, 250, and 500 mA. Reconstructions included both iterative reconstruction (IR) and standard FBP. The same vessels were examined using QCA (without simulated heart motion) in the projection showing the worst view of all stenoses as the reference standard.

CTA at 120 kVp, 100 kVp, 80 kVp and 500 mA, 250 mA and 100 mA showed very good agreement and correlation with QCA for coronary stenosis quantification. Tube current modulation did not have an overall significant effect on stenosis quantification with both reconstructions at all tube voltage levels (p>0.05, ANOVA). No significant difference was found between the limits of agreement comparing results with 120 kVp, 100 kVp and 80 kVp (p=0.05, F-test). Also no significant difference was found comparing the limits of agreement between IR and FBP with 120 kVp and 100 kVp (p>0.05 for both), with 80 kVp slightly narrower limits of agreement were shown with IR than with FBP (p=0.04). Through all stenoses no significant over- or underestimation was shown comparing manual stenosis quantification in CTA and QCA with both IR and FBP at 120 kVp, 100 kVp and 80 kVp (p>0.05 for all, t-test).  

In a moving heart phantom, 320-row CTA had very good agreement with QCA for stenosis quantification. Additionally using IR with adjusted scan parameters can further reduce effective radiation dose and may even improve diagnostic accuracy. 

These results indicate a potential clinical use of most recent generation CT scanners also for quantification of coronary stenosis and implicate the use of IR with adjusted scan parameters.

Kozma, P, Dewey, M, Quantification of Coronary Artery Stenosis with 320-Row CT Angiography in a Moving Heart Phantom: Comparison with Quantitative Coronary Angiography.  Radiological Society of North America 2014 Scientific Assembly and Annual Meeting, - ,Chicago IL. http://archive.rsna.org/2014/14045645.html