When someone has a stroke, the sooner they get appropriate medical attention, the better, and a new smartphone tool could help ensure that happens by allowing emergency personnel to determine whether a patient has indeed had a stroke.
Stroke symptoms can often be subtle or vague.
In these cases, patients must be evaluated with a series of tests upon arrival at the hospital. If these tests indicate that a stroke has occurred, after that The real treatment begins there, but if doctors know that an admitted patient is already a stroke patient, treatment can begin as soon as the patient arrives at the hospital.
That's where the experimental smartphone tool comes in.
Developed by a team of scientists at RMIT University in Australia, the software uses AI algorithms to analyze video of a patient's face as they smile, and if facial muscle movements are determined to be excessively asymmetric, the software alerts the user that the patient has recently suffered a stroke.
The system leverages existing facial action coding systems that break down facial expressions into individual components of muscle movements called “action units.”
“One of the key factors affecting stroke patients is that the facial muscles are usually unilateral, causing one side of the face to move differently than the other,” says lead scientist and doctoral student Guilherme Camargo de Oliveira. “We have AI and image processing tools that allow us to detect if there is a change in the asymmetry of the smile, which is key to detection in our case.”
When tested on facial videos of post-stroke patients and healthy volunteers, the system proved to be 82% accurate in identifying stroke patients, a figure that should improve as the technology is further developed.
The plan now is to turn this tool into an app that emergency personnel can use to detect strokes on the spot in a matter of seconds. In fact, scientists from the Polytechnic University of Valencia and Pennsylvania State University have already developed their own similar apps. In any case, it should be noted that more comprehensive tests, such as CT scans, will ultimately be carried out in the hospital.
A paper on this research was recently published in the journal. Computer methods and programs in biomedical sciences.
Source: RMIT University, via EurekAlert
