Sign In

Communications of the ACM

ACM TechNews

AI May Help Identify Patients with Early-Stage Dementia


View as: Print Mobile App Share: Send by email Share on reddit Share on StumbleUpon Share on Hacker News Share on Tweeter Share on Facebook

Researchers are studying whether artificial intelligence tools can be used to help clinicians better identify patients with early-stage dementia.

Credit: Philippe Lopez/Agence France-Presse/Getty Images

Scientists are exploring whether artificial intelligence can help clinicians better identify patients with early-stage dementia, with researchers and companies testing machine learning (ML) for mining and extracting meaning from vast, diverse datasets.

Much research involves sifting through electronic-health records to determine what combination of risk factors most accurately reflects cognitive decline.

A U.S. National Institute on Aging-funded study analyzed electronic records of more than 16,000 medical visits of 4,330 participants in a Kaiser Permanente Washington health system.

The researchers' model identified 31 factors associated with cognitive decline, flagging more than 1,000 visits leading to a dementia diagnosis.

Other studies have gauged written speech patterns for indicators of mental deterioration, and distinguished between cognitively impaired and healthy individuals; ML also has been used to forecast a patient's long-term dementia outcome, based on certain disease biomarkers.

From The Wall Street Journal
View Full Article - May Require Paid Subscription

 

Abstracts Copyright © 2020 SmithBucklin, Washington, DC, USA


 

No entries found