2023 ARRS ANNUAL MEETING - ABSTRACTS

RETURN TO ABSTRACT LISTING


2432. Virtual Monoenergetic Reconstructions Influence the Accuracy of Stenosis Measurements of Coronary CTA Using A Photon Counting CT
Authors * Denotes Presenting Author
  1. Tilman Emrich; Medical University of South Carolina; University Medical Center
  2. Elias Wolf; Medical University of South Carolina; University Medical Center
  3. U. Joseph Schoepf; Medical University of South Carolina
  4. Emese Zsarnoczay; Medical University of South Carolina
  5. Nicola Fink *; Medical University of South Carolina
  6. Moritz Halfmann; University Medical Center
  7. Akos Varga-Szemes; Medical University of South Carolina
Objective:
Coronary CT angiography (CCTA) has emerged as a clinical first-line test for the assessment of stable coronary artery disease (CAD). Conventional energy integrating systems suffer from certain limitations, including the overestimation of stenosis due to calcium blooming. With the recent introduction of clinical PCD-CT systems, the possibility to reconstruct virtual monoenergetic reconstructions (VMI) at different keV levels at high temporal resolution needed for cardiac imaging became available. The purpose of this study was to evaluate the influence of virtual monoenergetic reconstructions on the accuracy of coronary artery stenosis measurements on a clinical first-generation dual-source photon-counting detector computed tomography (PCD-CT) system, compared to invasive coronary angiography.

Materials and Methods:
Thirty-six patients (25 [69.4%] men, mean age 70.0 ± 9.1 years) underwent PCD-CCTA and invasive coronary angiography as part of their clinical workup for evaluation of stable CAD. Different keV levels from 40 - 140 keV in 10-keV steps were reconstructed with a slice thickness of 0.6 mm, Qr40 Kernel, and QIR strength level of 3. CT-based stenosis quantification of 62 calcified, mixed or soft tissue lesions was compared to quantitative coronary angiography (QCA).

Results:
Different keV reconstructions lead to significantly different stenosis measurements (e.g., mean CT-based stenosis: 67.5 ± 8.9% at 40 keV versus 53.9 ± 9.7% at 140 keV, p < 0.001). The optimal VMI keV level depended on the tissue quality of the plaque. Compared to the reference, all CT-based stenosis measurements led to an overestimation of stenosis by 5 - 20%.

Conclusion:
VMI reconstructions have the potential to improve the accuracy of stenosis quantification for PCD-CCTA. These spectral reconstructions may be used to increase the positive predictive value of CCTA, therefore potentially reducing the amount of unneeded invasive coronary angiographies.