Angiographic reconstructions of CT perfusion for occlusion detection in ischemic stroke

Abstract

BackgroundBoth CT angiography and CT perfusion involve the administration and tracking of a contrast medium bolus for different purposes. In this study, we aim to compare the diagnostic accuracy and subjective image quality of CTP-angiographic reconstructions with conventional CTA for occlusion detection in ischemic stroke patients. MethodsIn this retrospective study, patients with a final diagnosis of ischemic stroke and who underwent both CTA and CTP from September 2020 up to and including September 2021 were included. CTP-AR was reconstructed from the 1 mm CTP series at the time of maximum arterial inflow. Three readers with different levels of experience assessed both CTA and CTP for occlusion detection, expressed certainty, and rated subjective image quality. The reference standard was set at the consensus meeting. Pooled sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) were calculated. Differences in subjective and objective image quality and certainty were assessed using Kendall’s tau correlation and paired samples t-tests. ResultsIn total, 107/210 included patients had an occlusion based on our reference standard. Pooled sensitivity and specificity for occlusion detection were 90 % (95 %CI 85–93 %) and 94 % (95 %CI 90–97 %) for CTA, and 89 % (95 %CI 84–93 %) and 93 % (95 %CI 89–96) for CTP-AR, respectively. The pooled certainty did not significantly differ between CTA and CTP-AR (P=.43). The pooled subjective image quality scores significantly differed between CTA and CTP-AR (τ = 0.3, P < 0.001), where one reader rated subjective image quality higher in CTA and two readers in CTP-AR. ConclusionCTP-AR has comparable diagnostic accuracy to conventional CTA imaging for occlusion detection in ischemic stroke patients. Thus, using CTP-AR instead of CTA plus CTP may conserve contrast medium.