Combined application of multi-parameter semiquantitative Alberta stroke program early CT score to assess infarction severity in acute ischemic stroke

Abstract

Objective: To explore the significance of multi-parameter semiquantitative Alberta Stroke Program Early CT Score (ASPECTS) in detecting core infarction in acute ischemic stroke (AIS) patients. Methods: Twenty-one consecutive AIS patients from Beijing Hospital were retrospectively reviewed in this study from August 2016 to August 2017. All the patients presented within 6 hours since symptom onset.All the patients underwent one-stop dynamic whole brain 4D CTA-CTP scan before treatments (intravenous thrombolysis, mechanical thrombectomy, or other endovascular recanalization). MRI scan was performed in all patients 3 to 7 days after initial administration. Multi-parameter ASPECTS was calculated on the affected hemisphere regarding noncontrast CT (NCCT) map, cerebral blood flow (CBF) map, cerebral blood volume (CBV) map, mean transit time (MTT) map and MRI-DWI map by subtracting 1 point from 10 for any abnormalities visually detected. The evaluative consistency and diagnostic efficiency were analyzed by Pearson's χ(2) test, kappa identity test, and Kendall's coefficient of concordance using IBM SPSS statistics 22.0 software. Results: Compared with DWI-ASPECTS or follow-up NCCT-ASPECTS, CBV-ASPECTS had the best performance on both the kappa identity test (kappa coefficient κ=0.74)and Kendall's coefficient of concordance (τ=0.70). All the values of sensitivity (77.8%), specificity (95.5%), positive likelihood ratio (17.27) and AUC(0.87)were the highest in CBV-ASPECTS, followed by dMTT-ASPECTS. Meanwhile, the misdiagnosis rate (4.5%), missed diagnosis rate (22.2%) and negative likelihood ratio were the lowest in CBV-ASPECTS, followed by dMTT-ASPECTS. dMTT-ASPECTS had the same specificity(95.5%) as CBV-ASPECTS did and it could reveal the infarction in the areas where CBV map showed normal. NCCT-ASPECTS had the lowest sensitivity (64.7%) and highest missed diagnosis rate (35.3%). The misdiagnosis rate of CBF-ASPECTS was the highest (25.2%) and AUC was the lowest (0.76). Conclusions: Multi-parameter semiquantitative ASPECTS may predict the infarction accurately in AIS patients. Compared with the single parameter of NCCT-ASPECTS, this new method may have better detectability and diagnostic performance.