Automated versus manual imaging assessment of early ischemic changes in acute stroke: comparison of two software packages and expert consensus

Aim: This study aimed to compare the agreement between total ASPECTS and region-based scores from two automated ASPECTS software packages and an expert consensus (EC) reading with the final ASPECTS in patients who had prompt reperfusion from endovascular thrombectomy (EVT).

Methods: ASPECTS were retrospectively and blindly assessed by two software packages and EC on baseline non-contrast-enhanced CT (NCCT) images in patients with acute anterior circulation ischemic stroke with large vessel occlusion. Final ASPECTS from follow-up scans in patients with complete reperfusion within 100 minutes were compared to total and region-based scores.

Results: 52 patients met the criteria. Moderate agreement was observed between both software packages and EC for total ASPECTS, with no significant performance difference, though regional contributions differed: Package A was more sensitive in cortical areas but less specific, while EC and Package B were more sensitive for deep structures but less specific.

Conclusion: No clinically significant difference was found for total ASPECTS between human and automated packages, though regional scoring characteristics differed.