Related Experiment Video
Updated: Jul 11, 2025
07:26
Modeling Stroke in Mice: Transient Middle Cerebral Artery Occlusion via the External Carotid Artery
Published on: May 24, 2021
7.2K
StrokeClassifier: Ischemic Stroke Etiology Classification by Ensemble Consensus Modeling Using Electronic Health
Ho-Joon Lee1, Lee H Schwamm2,3, Lauren Sansing3
1Department of Genetics and Yale Center for Genome Analysis, Yale School of Medicine, New Haven, CT.
Research Square
|November 14, 2023
Summary
An artificial intelligence tool, StrokeClassifier, accurately predicts acute ischemic stroke causes using electronic health records. This AI tool shows potential for clinical decision support in stroke etiology determination.
Area of Science:
- Medical Informatics
- Artificial Intelligence in Medicine
- Neurology
Background:
- Determining the cause of acute ischemic stroke (AIS) is crucial for effective secondary prevention but remains diagnostically challenging.
- Existing methods rely on expert review of extensive patient data, which can be time-consuming and resource-intensive.
Approach:
- Developed and validated StrokeClassifier, an AI tool using natural language processing (NLP) on electronic health record (EHR) discharge summaries.
- The tool is an ensemble consensus meta-model of 9 machine learning classifiers trained on data from 2,039 AIS patients.
- Externally validated on the MIMIC-III dataset, achieving comparable performance to vascular neurologists.
Key Points:
- StrokeClassifier achieved a mean cross-validated accuracy of 0.74 and weighted F1 of 0.74 in predicting stroke etiology.
- Top predictive features included atrial fibrillation, age, and specific arterial occlusions (middle cerebral artery, internal carotid artery).
- A certainty heuristic reduced the proportion of cryptogenic strokes from 25.2% to 7.2% in a subset of patients.
Conclusions:
- StrokeClassifier demonstrates validated AI performance comparable to expert neurologists in classifying ischemic stroke etiology.
- The tool has potential as a clinical decision support system to aid in stroke diagnosis and management.
- Further development could enhance its utility in identifying stroke causes and refining secondary prevention strategies.
