A differential detailed diffusion-weighted imaging-ASPECTS for cerebral infarct volume measurement and outcome prediction

Abstract

Background: Diffusion-weighted imaging-Alberta Stroke Program Early Computed Tomography Score (DWI-ASPECTS) has been used to estimate infarct core volume in acute stroke. However, the same and indiscriminate score deduction for punctate or confluent DWI high-intensity lesion might lead to variation in performance. Aims: To develop and evaluate a differential detailed DWI-ASPECTS method in comparison with the conventional DWI-ASPECTS in core infarct volume measurement and clinical outcome prediction. Methods: We retrospectively recruited patients with acute ischemic stroke (AIS) treated with endovascular treatment between April 2013 and October 2019. In differential detailed DWI-ASPECTS, restricted diffusion lesion that was punctate or less than half of a cortical region (M1–M6) would not lead to subtraction of point. A favorable outcome was modified Rankin Scale score ⩽2 at 90 days after stroke onset. Results: Among 298 AIS patients, mean age was 75 years (interquartile range (IQR) 67–82), and 194 patients (65%) were males. Mean infarct core volume was 11 mL (IQR 3–37). Overall, the score by detailed DWI-ASPECTS was significantly higher than conventional DWI-ASPECTS (8 (7–9) vs. 7 (5–9); P < 0.01). The detailed DWI-ASPECTS resulted in a higher correlation coefficient (r) for core infarct volume estimation than the conventional DWI-ASPECTS (r = 0.832 vs. 0.773; P < 0.01). Upon re-classification of those scored ⩽6 in conventional DWI-ASPECTS (n = 134) by detailed DWI-ASPECTS, the rate of favorable outcome in patients with detailed DWI-ASPECTS >6 was significantly higher than those with ⩽6 (29 (48%) vs. 14 (19%); P < 0.01). Conclusions: Detailed DWI-ASPECTS appeared to provide a more accurate infarct core volume measurement and clinical outcome correlation than conventional DWI-ASPECTS among AIS patients treated with endovascular therapy.