First in-vivo coronary stent imaging with clinical ultra high resolution photon-counting CT

Abstract

Coronary stent imaging remains limited with conventional CT. In this patient study we evaluated the quality of coronary stent imaging and determined the optimal reconstruction settings for ultra-high-resolution (UHR) coronary CT angiography (cCTA) with clinical photon-counting-detector CT (PCD-CT). In this retrospective dual-center study, 22 patients with 36 coronary stents who underwent UHR cCTA with PCD-CT were included. Images with a slice thickness of 0.6mm and Bv40 kernel and UHR images at a slice thickness of 0.2mm with kernels of eight sharpness levels (Bv40, Bv44, Bv56, Bv60, Bv64, Bv72, Bv80, and Bv89) and adapted matrix-sizes and field-of-views were reconstructed. Image noise, contrast-to-noise-ratio (CNR), in-stent diameters, and differences of in-stent attenuation compared with adjacent segments were measured. Stent strut sharpness was quantified using data derived from line profiles. Subjective in-stent lumen visualization was rated by two blinded, independent readers. In-vitro stent diameters were taken as reference standard. At increasing kernel sharpness, CNR decreased, in-stent diameter increased (1.8​±​0.5mm for 0.6mm/Bv40 to 2.5​±​0.5mm for 0.2mm/Bv89), and stent strut sharpness increased. Differences of in-stent attenuation decreased from 0.6mm/Bv40 to 0.2 mm/Bv60-Bv80 kernels, being not different from zero for the latter kernels (p​>​0.05). Percentage (absolute) differences of measured to in-vitro diameters decreased from 40.1​±​11.1% (1.2​±​0.4mm) for 0.6mm/Bv40 to 16.6​±​8% (0.5​±​0.3mm) for 0.2mm/Bv89. There were no associations between stent angulation and in-stent diameter or attenuation differences (p​>​0.05). Qualitative scores increased from suboptimal/good for 0.6mm/Bv40 to very good/excellent for 0.2mm/Bv64 and 0.2mm/Bv72. UHR cCTA with clinical PCD-CT enables excellent in-vivo coronary stent lumen visualization.