Separating stroke and acute unilateral vestibulopathy using history, examination and vestibular tests: a machine learning approach

Acute unilateral vestibulopathy (AUVP) and posterior circulation stroke (PCS) are the two common causes of the acute vestibular syndrome. We developed and evaluated machine learning models to differentiate AUVP and PCS using combinations of patient history, examination, and vestibular tests.

Continue reading at PubMed →

If validated prospectively, a machine learning model combining history, examination, and vestibular test data could meaningfully improve stroke-vs-vestibulopathy triage, but clinicians should not change diagnostic protocols until external validation studies are published.

Misdiagnosing a posterior-circulation stroke as benign vestibular neuritis is a known patient-safety hazard; improved diagnostic algorithms could prevent missed strokes in dizziness presentations.

1. 01Machine learning models were trained to separate acute unilateral vestibulopathy (AVU) from posterior circulation stroke.
2. 02Inputs included patient history, neurological and vestibular examination findings, and vestibular function tests.
3. 03Automated differentiation could reduce missed stroke diagnoses in acute dizziness presentations.
4. 04This is an observational/model-development study; prospective external validation is needed.
5. 05Published in Journal of Neurology, a high-impact peer-reviewed neurology journal.

PMID

42709223

DOI

10.1007/s00415-026-13779-0.

Journal

Journal of Neurology

Publication type

research_article

Evidence level

2b

Population

Patients presenting with acute dizziness, including those with acute unilateral vestibulopathy and posterior circulation stroke

Intervention

Machine learning classification model using clinical history, examination, and vestibular test data

Comparator

Clinical diagnosis without machine learning