New study examines AI-driven evaluation for stroke patients

A team of experts developed an artificial neural network model to screen stroke patients, according to research out of Blacksburg-based Virginia Tech and Danville, Pa.-based Geisinger Health System.

The researchers — led by Vida Abedi, PhD, a Geisinger Health System researcher and an adjunct faculty member in the Biocomplexity Institute of Virginia Tech's Nutritional Immunology and Molecular Medicine Laboratory — examined EHR data from 260 emergency room patients experiencing stroke-like symptoms. They developed an ANN model to stratify patients exhibiting stroke symptoms.

Their findings, published in Stroke, sought to validate the model. The researchers demonstrated the model's average sensitivity for the diagnosis of acute cerebral ischemia was 80 percent. Its average specificity for diagnosis was 86.2 percent and its median precision for diagnosis was 92 percent. They concluded the model was an effective tool evaluate stroke patients.

"Our team has applied AI successfully to develop a data-driven triage process for classifying stroke patients," said Dr. Abedi.

More articles on health IT:

Regenstrief, IU establish informatics training program with $2.5M NIH grant

Nokia faces criticism over updated Withings 'health mate' app

USC to expand biotech community

Copyright © 2024 Becker's Healthcare. All Rights Reserved. Privacy Policy. Cookie Policy. Linking and Reprinting Policy.

 

Featured Whitepapers

Featured Webinars