Abstract

Abstract Funding Acknowledgements Type of funding sources: None. Background Spectral CT imaging and virtual non-contrast (VNC) reconstruction techniques allow coronary artery calcium score (CACS) calculations from coronary CT angiography (CCTA) datasets. A novel virtual material decomposition algorithm (PureCalcium, PC) developed for photon-counting detector CT (PCD-CT) yields the potential of further increasing the accuracy of CCTA-based CACS measurements, however, little is known about the means of optimizing its performance for routine Agatston-scoring. Purpose The aim of the current study was to determine the predictors of the performance of a novel virtual non-iodine algorithm for PCD-CT. Materials and Methods Calcium scoring on non-enhanced scans is predominantly used as a complementary examination to CCTA increasing both acquisition time and radiation dose, thus, acquiring accurate CACS from CCTA datasets are warranted. In the current study, CACS and CAC volumes were calculated using true non-contrast (TNC) images and PC reconstructions in 139 consecutive patients (61.5 ± 10.9 years, 41.7% female) undergoing clinically indicated CCTA. Relative differences (TNC–PC/TNC) of PC-based CACS calculations from non-enhanced scans was calculated for each patient. Predictors of the degree of discrepancy were determined using linear regression analysis. Results Overall, BMI and heart rate did not affect the accuracy of PC measurements. However, the application of a higher tube voltage (140 kVp instead of 120 kVp) significantly decreased the relative difference of PC-based measurements from TNC values (β=-0.33 [95%CI: −0.64 – −0.03]). Furthermore, the average volume/lesion ratio also proved to be an independent predictor of performance (β=-0.01 [95%CI: −0.02 – −0.004]). Conclusions Spectral imaging based PC reconstructions may substitute conventional calcium scoring for a selected population of patients. It seems plausible that its performance can be improved by increased tube voltage protocols, provided spectral information gathered from CTA datasets is enhanced by using 140 kVp. It must be noted that the accuracy of PC may be hindered for smaller calcified lesions.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call