Performance and influencing factors of using computed tomography perfusion to identify acute lacunar infarction: a retrospective single-center study

BackgroundLacunar infarction accounts for almost 25% of ischemic strokes.PurposeTo evaluate the effectiveness of computed tomography perfusion (CTP) in identifying acute lacunar infarction (ALI) by comparing its performance with non-contrast CT (NCCT) and CT angiography (CTA), and to study the potential influence of various imaging characteristics on detection accuracy.Material and MethodsA total of 309 patients who underwent baseline CT and follow-up diffusion-weighted imaging due to lacunar symptoms were enrolled. The detection performance of NCCT, CTA, CTP, and various CTP-derived parametric maps for identifying ALI was calculated and compared. In addition, the study examined and compared the performance of CTP across different subgroups, categorized based on infarction location, infarction size, and degree of white matter hyperintensity.ResultsALI was identified in 184 patients. CTP demonstrated significantly higher sensitivity (44.6%) in detecting ALI compared to NCCT (8.2%) and CTA (12.0%) (both <i>P</i> < 0.001). Among the four CTP-derived parametric maps, sensitivity ranged from 2.2% (cerebral blood volume [CBV]) to 41.8% (mean transit time [MTT]). In subgroup analyses, CTP showed higher sensitivity for detecting cortical lesions (60.0%) compared to posterior lesions (41.0%) (<i>P</i> = 0.061) and subcortical lesions (39.8%) (<i>P</i> = 0.035). CTP showed slightly higher sensitivity in detecting ALI with a larger infarct size (>10.6 mm) and in cases with mild-to-moderate WMH. However, these differences were not statistically significant (>10.6 mm vs. ≤10.6 mm, 45.2% vs. 44.0%, <i>P</i> = 0.870; mild-to-moderate vs. severe WMH: 45.3% vs. 41.2%, <i>P</i> = 0.660).ConclusionIn this retrospective study, we found that CTP outperformed NCCT and CTA in detecting ALI. CTP demonstrated higher sensitivity for detecting ALI compared to posterior and subcortical lesions.