Photon-Counting Detector CT Virtual Monoenergetic Images for Coronary Artery Stenosis Quantification: Phantom and In Vivo Evaluation

Elias V. Wolf, Moritz C. Halfmann, Akos Varga-Szemes, Nicola Fink, Roman Kloeckner, Stefanie Bockius, Thomas Allmendinger, Junia Hagenauer, Till Koehler, Karl Friedrich Kreitner, U. Joseph Schoepf, Thomas Munzel, Christoph Duber, Tommaso Gori, Yang Yang, Michaela M. Hell, Tilman Emrich*

*Corresponding author for this work
4 Citations (Scopus)

Abstract

BACKGROUND. Calcium blooming causes stenosis overestimation on coronary CTA. OBJECTIVE. The purpose of this article was to evaluate the impact of virtual monoenergetic imaging (VMI) reconstruction level on coronary artery stenosis quantification using photon-counting detector (PCD) CT. METHODS. A phantom containing two custom-made vessels (representing 25% and 50% stenosis) underwent PCD CT acquisitions without and with simulated cardiac motion. A retrospective analysis was performed of 33 patients (seven women, 26 men; mean age, 71.3 ± 9.0 [SD] years; 64 coronary artery stenoses) who underwent coronary CTA by PCD CT followed by invasive coronary angiography (ICA). Scans were reconstructed at nine VMI energy levels (40-140 keV). Percentage diameter stenosis (PDS) was measured, and bias was determined from the ground-truth stenosis percentage in the phantom and ICA-derived quantitative coronary angiography measurements in patients. Extent of blooming artifact was measured in the phantom and in calcified and mixed plaques in patients. RESULTS. In the phantom, PDS decreased for 25% stenosis from 59.9% (40 keV) to 13.4% (140 keV) and for 50% stenosis from 81.6% (40 keV) to 42.3% (140 keV). PDS showed lowest bias for 25% stenosis at 90 keV (bias, 1.4%) and for 50% stenosis at 100 keV (bias, -0.4%). Blooming artifacts decreased for 25% stenosis from 61.5% (40 keV) to 35.4% (140 keV) and for 50% stenosis from 82.7% (40 keV) to 52.1% (140 keV). In patients, PDS for calcified plaque decreased from 70.8% (40 keV) to 57.3% (140 keV), for mixed plaque decreased from 69.8% (40 keV) to 56.3% (140 keV), and for noncalcified plaque was 46.6% at 40 keV and 54.6% at 140 keV. PDS showed lowest bias for calcified plaque at 100 keV (bias, 17.2%), for mixed plaque at 140 keV (bias, 5.0%), and for noncalcified plaque at 40 keV (bias, -0.5%). Blooming artifacts decreased for calcified plaque from 78.4% (40 keV) to 48.6% (140 keV) and for mixed plaque from 73.1% (40 keV) to 44.7% (140 keV). CONCLUSION. For calcified and mixed plaque, stenosis severity measurements and blooming artifacts decreased at increasing VMI reconstruction levels.

Original languageEnglish
Article numbere2330481
JournalAmerican Journal of Roentgenology
Volume222
Issue number3
ISSN0361-803X
DOIs
Publication statusPublished - 03.2024

Fingerprint

Dive into the research topics of 'Photon-Counting Detector CT Virtual Monoenergetic Images for Coronary Artery Stenosis Quantification: Phantom and In Vivo Evaluation'. Together they form a unique fingerprint.

Cite this