Artificial Intelligence in Acute Neuroimaging Pathways: Diagnostic Accuracy, Workflow Performance, and Clinical Outcomes Across CT and MRI in Stroke and Trauma

This review summarizes current evidence on artificial intelligence (AI) applied to emergency neuroimaging for stroke and traumatic brain injury. Across diverse settings, most work focuses on CT-based tools that assist with detection, scoring, triage, and outcome prediction, with MRI used less often for complementary tasks. Findings generally suggest improved diagnostic support and early signals of workflow benefit, though study designs and reporting are heterogeneous. Overall, AI appears ready to augment acute care when integrated into decision pathways, but broader, practice-oriented evaluations with attention to robustness, equity, and real-world implementation remain necessary.