I put my face on my smartphone… “I’m having a stroke, get to the hospital quickly.”

Early diagnosis by detecting facial asymmetry
Stroke treatment rates increase as emergency treatment progresses
Expectations for chronic disease and healthcare apps

A screenshot of the healthcare app. photo source procarement

An artificial intelligence (AI)-based smartphone application has been developed that almost accurately detects facial asymmetry, a major symptom of stroke. Healthcare apps, which have been used primarily for the correction of cognitive behavior and mental illnesses such as sleep disorders and addiction, are expanding their scope to include chronic and acute diseases.

A research team led by Professor Radoslav Raitsev from the Department of Neurology at the University of California, Los Angeles (UCLA) developed an AI-based smartphone app called ‘FAST.AI’. It is an app that captures stroke symptoms as accurately as a neurologist and allows you to receive appropriate treatment early. The results of the study will be presented at the 2023 International Stroke Conference to be held in Dallas, USA on the 8th to the 10th (local time).

Stroke is the fifth leading cause of death and one of the leading causes of disability in the United States. About 85% of strokes in the United States are ‘ischemic strokes’ where blood flow to the brain is blocked due to atherosclerosis of the blood vessels that supply blood to the brain. According to the American Stroke Association under the American Heart Association, in the case of an ischemic stroke, 1.9 million brain cells are lost per minute if not treated in time. After an hour, 120 million brain cells are lost. No matter how late it is, it must be treated between 3 and 4 and a half hours to save even some of the remaining brain cells. According to a previous study, those who received treatment within 90 minutes of the first symptoms appearing were three times more likely to recover than those who took more than 90 minutes.

FAST.AI uses machine learning algorithms to identify common stroke symptoms, such as facial asymmetry, sensory disturbances such as numbness or weakness in the arms, and language changes. It analyzes 68 specific points in the patient’s face image, and the smartphone sensor detects the movement and direction of the arm, and the voice recognition module detects changes in voice to find stroke symptoms.

The research team checked the performance of FAST.AI on approximately 270 patients diagnosed with acute stroke within 72 hours of hospitalization in four Bulgarian cities between July 2021 and July 2022. Their average age was 71 years, and 41 % are women. The analysis showed that FAST.AI detects stroke-related facial asymmetry in patients with almost 100% probability. The accuracy of identifying a sensory impairment in the arm was more than two-thirds. The voice recognition module needs further testing, but preliminary analysis confirmed that it can reliably detect patients’ slurred voices.

“In a situation where many stroke patients do not come to the hospital in an appropriate time for clot dissolving treatment, FAST.AI will help detect the signs and symptoms of stroke early,” said Professor Reicheff.

‘Reset’, a mobile app for drug addiction treatment developed by an American start-up in 2017, proved the effectiveness of cognitive-behavioral treatment for people addicted to alcohol, cannabis and cocaine and received approval as a new drug from the Ministry of Food and US Drug (FDA) for the first time. Since then, smartphone apps have begun to attract attention for prognosis management for diseases that have a major impact on improving lifestyle, such as cognitive behavioral therapy such as anxiety disorder, panic disorder, and drug addiction, and diabetes and hypertension . A typical example is the app for heart failure patients, ‘ProHerz’. It transmits bio-signals such as blood pressure, heart rate, blood oxygen saturation, weight, and body temperature to the Care Center system, and encourages behavior improvement by providing easy-to-use graphic feedback to the patient. There is also a car driving game app that improves the attention of attention deficit hyperactivity disorder (ADHD) patients, and an app that helps diabetics lose weight by managing their lifestyle.

Donga Science Reporter Yoon Young-hye yyh@donga.com

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.