Mobile app shows promise for real-time stroke detection

A smartphone-enabled machine learning algorithm may be as good as a neurologist at identifying signs and symptoms of acute stroke, according to preliminary research presented at the International Stroke Conference.

“Many stroke patients don’t make it to the hospital in time for anti-clot treatment, which is one of the reasons why it’s essential to recognize stroke symptoms and call 9-1-1 immediately. Radoslav I. Raychev, MD, FAHA, clinical professor of neurology at the University of California, Los Angeles, said in a related press release.

Raychev and his colleagues developed FAST.AI, a smartphone app designed to recognize strokes using machine learning algorithms, which identified typical symptoms such as facial asymmetry, limb weakness superiors and speech impairments.

The researchers analyzed data from 269 people diagnosed with acute stroke (median age, 71 years; 41% female) who were admitted to four major metropolitan stroke centers in Eastern Europe between July 2021 and July 2022. Data capture of speech patterns and facial expressions occurred via video recording, with data on the arms collected by sensors on the device.

Algorithm elements include 68 facial landmarks to measure asymmetry, an agnostic classifier to detect arm weakness, and a frequency analysis component to detect abnormal or garbled speech. The researchers performed all the tests within 72 hours of admission and compared each machine learning output with the neurologists’ clinical impression.

According to the results, analyzes of 18,311 facial images demonstrated a significant degree of sensitivity (99.42%), specificity (93.67%) and accuracy (97.11%) in detecting facial asymmetrywhile results from 43 motion trajectories detected arm weakness with a sensitivity of 71.42%, a specificity of 72.41% and an accuracy of 72.09%.

The researchers also reported that preliminary analysis of the speech alteration algorithms confirmed adequate features to detect anomalies.

“The initial results confirm that the app reliably identified symptoms of acute stroke as accurately as a neurologist, and they will help improve the app’s accuracy in detecting signs and symptoms of stroke. a stroke,” Raychev said in the statement.

Reference:

ADD A SUBJECT TO EMAIL ALERTS

Receive an email when new articles are published on

Please provide your email address to receive an email when new articles are published on .

” data-action=subscribe> Subscribe

We have not been able to process your request. Please try again later. If you continue to have this problem, please contact [email protected].

Return to Healio.

International Stroke Conference