Skip to main content
Algorithms analyze EHR data to predict Alzheimer's
2/13/2020

Two different machine learning algorithms described in Artificial Intelligence in Medicine and the Journal of the American Geriatrics Society accurately identified patients with Alzheimer's disease and related dementia from structured and unstructured EHR data, providing a scalable, inexpensive screening method, says lead researcher Malaz Boustani. Both a natural language processing algorithm and a random forest model based on decision trees accurately predicted the onset of dementia, and medical notes proved to be the most valuable data source for identifying high-risk patients.

Full Story: