Scientists at the University of California Los Angeles use rapid diagnosis with the help of Athletes
A new app by researchers at the University of California Los Angeles (https://www.ucla.edu) uses machine learning and artificial intelligence (AI) to detect stroke symptoms and quickly warn those at risk. Together with several medical institutions in Bulgaria, data from 240 stroke patients was used to train the app. Within 72 hours of the onset of symptoms, the researchers created videos of the patients using smartphones to detect facial asymmetry and speech changes in the patients.
Sound waves become images
The app measures facial changes using 68 landmarks. In addition, the researchers tested the strength of the arms using the device’s accelerometers. To determine speech changes, the experts use a method that translates sound waves into images to compare normal and slurred speech patterns. The developers were then able to test the app using neurologist reports and brain scan data, finding that the app is sensitive and specific enough to accurately diagnose a stroke in almost all cases.
“A quick and accurate assessment of symptoms is essential to ensure that people with stroke survive and regain their independence. We hope that the use of this app will significantly improve stroke care,” clarifies Radoslav Raychev, vascular and interventional neurologist at US University.
One million people affected in Europe
The researchers presented their app at the 20th annual meeting of the Society of NeuroInterventional Surgery (https://www.snisonline.org/meetings/snis-20th-annual-meeting-and-fellows-course/), which just ended in San Diego. The app is primarily intended for emergency paramedics and emergency doctors, who can use it to make a quick diagnosis even if they do not have detailed expertise in the field.
In the European Union alone, more than one million people suffer a stroke every year. In this case, the brain is suddenly no longer supplied with sufficient blood due to the formation of clots. The faster the clot is dissolved, the greater the patient’s chances of surviving or even recovering completely or to a large extent.
Ethanol-enhanced nanosensor: production also more environmentally friendly (Photo: mq.edu.au)