Abstract
To develop and validate a clinical-laboratory screening model for rapid differentiation of intracranial aneurysm (IA)-related acute ocular motor nerve palsy (AOMNP) from microvascular ischemia (MI), thereby reducing diagnostic delays and optimizing resource utilization. In this multicenter retrospective study, 62 AOMNP patients from Anhui Provincial Hospital (derivation cohort: IA = 29, MI = 33) and 29 from Suzhou Municipal Hospital (validation cohort: IA = 9, MI = 20) were categorized based on rigorous imaging criteria. Independent predictors of IA were identified via multivariable logistic regression, followed by nomogram construction. Diagnostic performance was evaluated using receiver operating characteristic (ROC) curves, with model generalizability assessed through external validation. Female sex (adjusted odds ratio [OR] = 7.94, 95% CI: 1.60-40.00), pupillary changes (OR = 10.02, 95% CI: 2.22-45.15), and elevated D-dimer levels (> 0.675 mg/L; OR = 1.58 per 0.1 mg/L increment, 95% CI: 1.05-2.39) emerged as independent predictors of IA. The integrated model demonstrated robust discriminative capacity (AUC = 0.899, derivation cohort), with preserved performance in the validation cohort (sensitivity: 77.8%, specificity: 85.0%, negative predictive value: 89.5%). Implementation of a simplified triage algorithm (female sex + pupillary abnormalities + D-dimer > 0.675 mg/L) prioritized high-risk patients for urgent vascular imaging, potentially reducing unnecessary testing by 73.3% in low-risk subgroups.