医学
回顾性队列研究
核医学
可视化
放射科
探测器
光子计数
队列研究
前瞻性队列研究
医学物理学
多中心研究
双盲研究
作者
Ilse Huijberts,Bart A.J.M. Wagemans,Florentina M.E. Pinckaers,M M Q Robbe,Wim H. van Zwam,Robert J. van Oostenbrugge,Alida A. Postma
标识
DOI:10.1016/j.ejrad.2025.112618
摘要
PURPOSE: The visualisation of lenticulostriate arteries (LSA) using CT-techniques has not been extensively studied. Currently, digital subtraction angiography (DSA) is the preferred imaging modality for LSA visualisation. The new photon-counting detector CT (PCD-CT) may offer a non-invasive alternative. This study evaluates PCD-CT for LSA visualisation compared to DSA and energy-integrating detector CT-angiography (EID-CTA). METHODS: We assessed imaging data from subarachnoid haemorrhage (SAH) patients, as they are expected to have intact LSA and routinely undergo DSA. From 2021 to 2024, adult SAH-patients who underwent both DSA and either EID-CTA or PCD-CT were enrolled. PCD-CT was acquired in UHR mode. For LSA > 5 mm from the middle cerebral artery (MCA) origin, only the length of the largest branch was measured, whereas for LSA < 5 mm all individual branches were assessed. LSA presence, length and subjective image quality were assessed across PCD-CT, EID-CTA, and DSA by two radiologists. Cohen's kappa and the Intraclass Correlation Coefficient were used to evaluate inter-rater agreement. RESULTS: 159 patients were included. No differences were found between DSA and PCD-CT in LSA detection and length < 5 mm from the MCA origin. At > 5 mm from the MCA origin, the measured length of LSA was significantly longer on DSA. PCD-CT outperformed EID-CTA for LSA detection and length measurements. Observer 1 rated PCD-CT quality lower than DSA, while observer 2 found no difference. EID-CTA quality was rated significantly lower than both PCD-CT and DSA. Inter-rater agreement ranged from poor to moderate across modalities. CONCLUSION: PCD-CT holds promise as a method for non-invasive LSA visualisation and requires further evaluation in prospective studies and clinical settings.
科研通智能强力驱动
Strongly Powered by AbleSci AI