Artificial intelligence (AI) is considered a promising tool to improve stroke diagnosis in the ED. “Our team is leading the development of such models to be used in real time in clinical settings,” reports Ramin Zand, MD, an associate professor of medicine and neurology at Geisinger Health System in Danville, PA.
Strokes, especially posterior circulation strokes, are associated with significant diagnostic error in the ED.1 “Mainly, this is because some of the signs and symptoms may be misleading, and the ED provider may not request a consult with neurology,” Zand explains.
Machine learning models can be designed to capture subtle signs of stroke and assist ED providers in catching stroke patients who might otherwise go undetected. Regulatory agencies, including the FDA, are starting to guide the creation of AI systems.2 Still, the best standard of an AI-driven triage system for stroke must go further, according to Zand.
The authors of a recent paper offer a framework for a decision support system using AI and clinical data, in combination with patients’ presenting symptoms, to support ED providers in diagnosing stroke.3 As end-users of the tool, ED providers’ contribution is “invaluable. As a matter of fact, a successful implementation would not be possible without having ED providers’ inputs and direct involvement,” says Zand, one of the paper’s authors.
The goal is for AI models to “give ED providers a better, more comprehensive picture of the patient to empower them to make more informed decisions,” Zand adds.
REFERENCES
- Tarnutzer AA, Lee SH, Robinson KA, et al. ED misdiagnosis of cerebrovascular events in the era of modern neuroimaging: A meta-analysis. Neurology 2017;88:1468-1477.
- U.S. Food & Drug Administration. Proposed regulatory framework for modifications to artificial intelligence/machine learning (AI/ML)-based software as a medical device [SaMD]: Discussion paper and request for feedback. 2019.
- Abedi V, Khan A, Chaudhary D, et al. Using artificial intelligence for improving stroke diagnosis in emergency departments: A practical framework. Ther Adv Neurol Disord 2020;13:1756286420938962.
Strokes, especially posterior circulation events, are associated with significant diagnostic error in the ED. Machine learning models can be designed to capture subtle signs and assist providers in catching cases that might otherwise go undetected.
You have reached your article limit for the month. Subscribe now to access this article plus other member-only content.
- Award-winning Medical Content
- Latest Advances & Development in Medicine
- Unbiased Content