Impact of Photon-counting CT–based Virtual Monoenergetic Imaging on Detecting Myocardial Late Iodine Enhancement Compared to Cardiac MRI

Purpose: To investigate the image quality and diagnostic performance of photon-counting detector CT–based delayed enhancement (DE) at different virtual monoenergetic imaging (VMI) levels in detecting and characterizing myocardial fibrosis, using MRI as the reference. Materials and Methods: This post hoc subanalysis of a prospective research cohort included participants who underwent late gadolinium enhancement MRI between July 2021 and January 2022 and same-day research CT DE on a photon-counting detector CT system. CT DE scans at different VMI levels (40–90 keV) were evaluated by two blinded readers in terms of subjective and objective image quality and presence and distribution of myocardial fibrosis. Diagnostic performance measures were calculated, and agreement with MRI was evaluated with Cohen κ statistics. Results: The CT DE scans of 27 participants (mean age, 52.9 years ± 17.2 [SD]; 14 female) were analyzed. Twelve of 27 (44%) participants and 87 of 459 (19%) segments had a positive late gadolinium enhancement MRI. VMI at 50 keV demonstrated an adequate trade-off between objective (image noise and contrast-to-noise ratio) and subjective (image noise and sharpness and overall quality) image quality, exhibiting the highest sensitivity, specificity, and accuracy in focal fibrosis detection, in per-segment analyses (87.4% [95% CI: 80.4, 94.3], 97.8% [95% CI: 96.4, 99.3], and 95.9% [95% CI: 94.0, 97.7], respectively), with a strong agreement with MRI (κ=0.86). These reconstructions showed the highest concordance in discriminating fibrosis patterns. Conclusion: Photon-counting detector myocardial CT DE with VMI reconstructions at 50 keV represents a valuable tool for myocardial fibrosis detection and characterization, with best diagnostic performance and agreement with MRI when compared with other kiloelectron volt levels.