Favorable AI-based infarct core volume (≤35 mL) on noncontrast CT was independently associated with favorable 3-month outcome compared to unfavorable volume (OR 3.03; 95% CI 1.56–5.88; p=0.001).
Observational (n=290)
Yes
Does favorable AI-ICV on NCCT predict favorable clinical outcomes in acute ischemic stroke patients treated with reperfusion therapies?
Favorable AI-based infarct core volume estimation on noncontrast CT independently predicts favorable 3-month outcomes in acute ischemic stroke patients undergoing reperfusion therapy.
Effect estimate: OR 3.03 (95% CI 1.56-5.88)
Absolute Event Rate: 63.9% vs 36%
p-value: p=0.001
Abstract Background and aims CTP has been used to guide reperfusion selection in late-window stroke. Recently, AI-based infarct core volume estimation on NCCT (AI-ICV, Methinks) has been associated to clinical outcomes. We investigated the association between AI-ICV, CTP-patterns, and outcomes in reperfused stroke patients. Methods Consecutive patients undergoing admission NCCT+CTP and treated with reperfusion-therapies (EVT and/or IVT) were collected from a prospective registry across three stroke-centers (January 2021–May 2024). Admission AI-ICV was categorized as favorable (≤35 mL) or unfavorable (35 mL) based on prior observations. CTP-patterns were automatically determined (rCBF30%; RAPID/IschemiaView) and categorized per EXTEND criteria as favorable or unfavorable. Late-window thrombolysis was defined as symptom onset–to–arrival 4hours. Outcomes were favorable outcome (mRS 0–2 at 3-months) and symptomatic-intracranial-hemorrhage(sICH) Results From 290 patients included (mean age 71.8 SD±14.6 years; median NIHSS 10IQR:5–18); 49% presented in late-window. Primary-EVT was performed in 54.1%, and 20% received IV-thrombolysis only. Favorable outcome occurred in 54.1% Overall, favorable AI-ICV was associated with favorable outcome (63.9% vs 36.0%, p0.001) and remained independently associated after adjustment for CTP-pattern, age, and baseline NIHSS (OR 3.03, 95% CI 1.56–5.88; p=0.001, Figure1). Similar associations were observed in late-window (OR:3.91, 95% CI:1.23–12.47; p=0.021, Figure2). No association was observed between sICH and either AI-ICV or CTP-pattern. Conclusions Among reperfusion treated patients with available CTP, favorable AI-ICV on NCCT independently predicted favorable outcome, including in late-window, These finding support further evaluation of AI-ICV-based selection for reperfusion therapies, particularly late-window thrombolysis Conflict of interest Dr Ortega-Gutierrez: Grants: NIH-NINDS (R01NS127114-01, R03NS126804), PCORI, Stryker, Medtronics, Microvention, IschemiaView, Viz.ai, and Siemens. Consultant: Medtronic and Stryker Neurovascular. Dr Ribó is the Chief Medical Office of Methinks and has equity interests. Consultant: AptaTargets, Cerenovus, Medtronic, Philips, Stryker. Stock: Anaconda, Nora Figure 1 - belongs to Results Figure 2 - belongs to Results
Flores et al. (2026) conducted an observational in Acute ischemic stroke (n=290). Favorable AI-based infarct core volume (≤35 mL) vs. Unfavorable AI-based infarct core volume (>35 mL) was evaluated on Favorable outcome (mRS 0–2 at 3-months) (OR 3.03, 95% CI 1.56-5.88, p=0.001). Favorable AI-based infarct core volume (≤35 mL) on noncontrast CT was independently associated with favorable 3-month outcome compared to unfavorable volume (OR 3.03; 95% CI 1.56–5.88; p=0.001).