Mixed coronary plaque phantom analysis by photon-counting CT: impact of calcium and iodine on low-attenuation plaque...

To test whether photon-counting detector CT (PCD-CT) improves low-attenuation plaque (LAP) detection compared with energy-integrating detector CT (EID-CT) using a mixed-plaque phantom and full-rotation helical images as high-resolution reference.

We imaged a custom phantom housing three different sized mixed-plaque inserts (LAP_5.5mm, LAP_4mm, and LAP_7mm) combining iodinated lumen (600/1000 HU), calcified arcs (200/800 mg/cc), and a narrow LAP central strip (1-1.5 mm, 75 HU), embedded in a fat-equivalent ring (-60 HU). Polychromatic and virtual monoenergetic images (40/70/100/130 keV) were reconstructed with soft (Qr40) and sharp (Qr72) kernels and quantum iterative reconstruction (QIR) levels. Findings were compared with a clinically used EID-CT protocol. HU line profiles and edge-width-at-half-maximum (EWHM) at the iodine-phantom interface were measured.Reference EWHM was 0.125 mm. Cardiac half-rotation Qr40 reconstructions yielded EWHM 0.68-0.75 mm, whereas Qr72 approached reference values (0.18-0.23 mm). In LAP_5.5mm, the 1-mm LAP layer was detectable on standard and UHR images with sharp kernels and higher QIR, with best attenuation fidelity using Qr72. For the smallest LAP_4mm, only Qr72/QIR2 resolved the 1-mm LAP component, although measured attenuation remained elevated because of blooming from adjacent iodine and calcium. In LAP_7mm, Qr72/QIR4 most closely reproduced the nominal LAP attenuation. Lower-keV images (40-70) increased LAP masking from iodine. EID-CT did not clearly depict LAP_4mm.

PCD-CT provides superior spatial resolution and HU fidelity for detecting LAP, particularly when using UHR images combined with sharp kernels (Qr72) and higher iterative strength (QIR 2-4). LAP detection improved at 130 keV vs. lower energies.