Multilayer and crushed stent visualization using photon- counting detector computed tomography - a proof of concept

Photon‑counting detector CT (PCD‑CT) enables ultra‑high resolution (UHR, 0.2 mm) and may improve visualization of coronary stents compared with standard‑resolution (SR, 0.4 mm). In this ex- vivo study, we utilized a phantom heart model to simulate physiological attenuation and imitated the most common coronary bifurcation stenting techniques by creating multiple stent layers and stent- crush situations (single layer, crushed, two‑ and three‑layer). We compared UHR and SR PCD‑CT using Bv72 and Bv56 kernels. Objective endpoints were in‑stent lumen visibility, signal-to-noise ratio, and percentage change in stent attenuation; subjective image quality was rated on a 5‑point Likert scale (1 = excellent to 5 = non‑diagnostic). Depending on distribution, two‑sided t‑tests (mean ± SD) or Mann-Whitney U tests (median[IQR]) were applied, p‑values were Bonferroni‑corrected for multiple comparisons. In the pooled Crush cohort, UHR Bv72 improved in‑stent lumen visibility when compared with both, SR Bv72 (single-layer: 68.08 ± 4.30% vs. 79.03 ± 3.79%; crushed: 63.62 ± 4.97% vs. 74.73 ± 2.51%) and SR Bv56 (single-layer: 63.63 ± 4.47% vs. 79.03 ± 3.79%; crushed: 62.18 ± 3.55% vs. 74.73 ± 2.51%), all p < 0.01. The effect was even more pronounced in small stents (single‑layer 65.45 ± 1.91% (SR) vs. 77.11 ± 2.55% (UHR) and crushed 60.67 ± 2.67% vs. 74.00 ± 2.67%), both p < 0.01. Regarding subjective image quality, UHR was associated with better sharpness and less blooming, e.g., UHR Bv72 vs. SR Bv56 (multilayer: sharpness 1[0] vs. 2[0], blooming 1[0] vs. 2[0]; crush: sharpness 1[0] vs. 2[0], blooming 1[0] vs. 2[1]; all p ≤ 0.01) and UHR Bv72 vs. SR Bv72 (multilayer: sharpness 1[0] vs. 1[1], blooming 1[0] vs. 1[1]; crush: sharpness 1[0] vs. 1[1]; all p ≤ 0.05). These findings indicate that UHR PCD‑CT with Bv72 might enable assessment of complex stents in a phantom setting; even though clinical studies are warranted.