Biomedical Engineers at RMIT University have developed an AI-powered mobile application that could help doctors detect early signs of stroke within seconds.
The revolutionary software, which RMIT University researchers developed in collaboration with Brazil’s São Paulo State University, can immediately detect stroke by quickly analyzing facial expressions, including facial symmetry and muscle movements, raising hope for early stroke detection and improved treatment outcomes.
According to the research findings published in April 2024 in the journal Computer Methods and Programs in Biomedicine, the AI-backed tool has achieved an accuracy of 82% based on Kiss and Spread, with a sensitivity of 91% in detecting stroke by analyzing “subtle and specific facial movements and are interpretable to the facial expressions.”
This technology could be game-changing, especially in early stroke detection, with the potential for computerized video assessment of patients by fast responders to initiate the treatment process as soon as possible.
According to the researchers, the AI-powered tool features action units that can read and interpret subtle and specific facial movements associated with post-stroke, leading to early treatment and improved patient outcomes.
Revolutionary Stroke Detection Technology— Why It’s Important
Stroke is the leading cause of disability and the second leading cause of death worldwide, according to the World Health Organization (WHO), with one in four people aged over 25 experiencing stroke in their lifetime, according to the World Stroke Organization. These reports show how devastating a stroke can be to anyone.
Considering that the conventional method of stroke detection diagnosis involves complex, costly, and time-consuming methods, many people do not have access to life-changing early diagnosis before brain damage occurs, which in most cases turns fatal.
According to a 2024 report released by the American Heart Association, stroke accounts for 1 in every 21 deaths in the United States, with approximately 162,890 people reportedly dying from stroke in 2021 alone.
Due to its lethality, stroke has become one of the leading contributors to the healthcare burden, with about $252.2 billion reported in 2021, according to Heart Disease and Stroke Statistics—2022 Update.
More Testing of the New Technology
According to the researchers, the results exhibited by the cutting-edge smartphone tool would need extensive real-world testing to determine its effectiveness in stroke detection and potential for integration into the healthcare sector.
“This needs to be tested in real-world conditions, with people of different ethnicities and smartphone use,” researchers noted in the report, adding it could help first responders perform screening tests and initiate early treatments to improve patient outcomes.
Meanwhile, the researchers are seeking partners to help turn the new digital stroke diagnostics tool into a functional mobile app and explore its potential for detecting other types of neurological conditions.
