Introduction: Artificial intelligence-based (AI) software, like Brainomix e-ASPECTS, has gained attention for its potential in improving the evaluation of neuroimaging studies for acute ischemic stroke. This study aimed to assess whether Brainomix e-ASPECTS software improves interobserver agreement and accuracy in detecting ASPECTS regions affected by anterior circulation large vessel occlusion (LVO) compared to standard image interpretation without AI.

Methods: We included 60 consecutive patients with anterior circulation LVO who underwent TICI 3 revascularization within 60 minutes of baseline CT. A total of 16 readers (senior and junior neuroradiologists and vascular neurologists) participated. The readers first interpreted CT scans independently, assessing the ASPECTS and affected regions. Two months later, the same CT scans were re-evaluated with the assistance of e-ASPECTS software. Interclass correlation coefficient (ICC) for total ASPECTS and interobserver agreement (using Fleiss’ Kappa) were assessed with and without the AI assistance. Accuracy was evaluated by comparing the results to the 24-hour CT scans, which served as the ground truth.

Results: Without e-ASPECTS assistance, the ICC for total ASPECTS was 0.395 (fair agreement), while it increased to 0.574 (good agreement) with e-ASPECTS assistance (P < 0.01). Significant improvement in inter-rater agreement was observed with e-ASPECTS for each region, except for M6 and caudate. The software also showed higher accuracy in detecting affected regions compared to the overall cohort of readers (both with and without e-ASPECTS assistance), except for the caudate.

Conclusions: Brainomix e-ASPECTS software significantly improved inter-rater agreement and accuracy in evaluating ASPECTS scores. It outperformed the overall group of neuroradiologists and neurologists in predicting final infarct/ASPECTS, indicating its potential value for enhancing stroke image assessments.