Abstract Background and aims Distinguishing acute ischemic stroke (AIS) from stroke mimics (SM) remains challenging in emergency neurology, with up to one-third of suspected cases ultimately identified as mimics despite standardised pathways and rapid imaging. Misclassification may lead to unnecessary thrombolysis or delayed reperfusion. Unlike previous single-domain models, this study integrates bedside clinical and early imaging data from real-world acute stroke workflows. Methods This prospective multicentre study included 1,662 consecutive patients evaluated for suspected AIS (507 SM, 30.5%). An elastic-net logistic regression model was trained on 1,411 patients using routinely collected clinical variables (demographics, comorbidities, medications, NIHSS, FAST, FAST-Plus) and early imaging data (Brainomix e-Stroke v10.1). NCCT was classified as positive only for ischemic changes, CTA as positive for vessel occlusion or stenosis, and CTP as positive/negative with a “not performed” category, including discordant results. Model performance was assessed on an independent test set (n = 251). Results The combined model achieved strong discrimination (AUC 0.89; 95% CI: 0.85–0.93). Predictors of SM included negative NCCT and CTA, unperformed CTP, epilepsy, and headache at presentation, and missing NIHSS data. Positive NCCT, CTA, CTP, smoking, and antiplatelet therapy predicted AIS/TIA. At the optimal ROC threshold, sensitivity for AIS/TIA was 75.9% and specificity 83.1% (PPV 91.0%, NPV 60.4%). Conclusions A model integrating routine clinical and early imaging data reliably differentiates AIS/TIA from SM. By bridging bedside assessment and imaging confirmation, it supports safer and faster decision-making in acute stroke care. Supported by the MoH of the Czech Republic (NU23-04-00336), and by MEYS, LRI CZECRIN (LM2023049). Conflict of interest Katerina Dvornikova:nothing to disclose, Veronika Kunesova: nothing to disclose, Svatopluk Ostry: nothing to disclose, Petr Novobilsky: nothing to disclose, Lenka Bártová: nothing to disclose, Linda Machova: nothing to disclose, Beata Stanková:nothing to disclose, István Szegedi:nothing to disclose, Adéla Kondé:nothing to disclose, Jana Licha:nothing to disclose, Kristyna Janotova:nothing to disclose, Martin Reiser:nothing to disclose, Tomas Jonszta:nothing to disclose, Jaroslav Havelka:nothing to disclose, Ondrej Volny: nothing to disclose, Michal Bar1:nothing to disclose
Dvorníková et al. (Fri,) studied this question.