Image Based Automated ASPECT Score for Acute Ischemic Stroke Patients

Abstract

The Alberta Stroke Program Early Computed Tomography Score (ASPECTS) is a tool to assess early ischemic changes in acute stroke patients and found to be superior when applied to contrast-based image modalities. We hypothesized that automated ASPECTS scores have no differences with the manual scoring. We generated time-invariant CTA (tiCTA) from CT Perfusion dataset and measured the ASPECTS score automatically and manually. Statistical analysis was performed to see the differences. The association of both measurements with patient outcome was measured using NIH stroke scale at admission, final infarct size, and 3-month modified Rank Scale (mRS) determined using the Spearman correlation coefficient. As a result, the difference between automated and manual ASPECT scores was statistically not significant (p=0.18). Both automated ASPECTS scores were identical in 40% of the patients for the total score. While all ASPECTS scores have limited association with outcome, our study illustrates the usability of automated ASPECTS applied on tiCTA, allowing simplification of CT workflow for acute ischemic stroke patients and promoting faster analysis.