In Australia, stroke is one among the leading causes of death and everlasting disability. 5% of deaths are attributable to strokes, while strokes cost the Australian health system 6.2 billion Australian dollars yearly.
Strokes occur when there’s a sudden lack of blood supply within the brainThis deprives brain tissue of the oxygen and nutrients it needs, which may lead to break to parts of the brain.
Timely stroke treatment can limit brain damage and improve the prognosis for patients. However, this requires early detection of symptoms, which will not be at all times easy.
Our team has developed a brand new smartphone app that examines an individual’s facial expressions and detects whether or not they have suffered a stroke. We recently published the outcomes of a Pilot study this tool and located that it may possibly quickly and comparatively accurately determine whether someone has had a stroke.
Scan facial expressions
The earliest outward symptoms of a stroke include facial expressions akin to a drooping face, where one side of the mouth will not be activated when an individual tries to smile.
But paramedics who reply to emergencies and Hospital emergency room Strokes often go unnoticed by staff. Facial expressions vary from individual to individual, and it’s difficult to detect subtle changes in a stressful environment. This may be even harder if the patient is from a special ethnic group or cultural background.
Using our smartphone app, a paramedic or other first responder asks the patient to smile while “filming” the patient's face. An AI-based model then analyzes the video recording and appears for similar signs that doctors use to detect a stroke, namely asymmetrical drooping of the mouth.
Seamus Daniel, RMIT University
The app is designed for simplicity – the user just must point the camera on the patient and press a button. To protect the patient's privacy, the video is analyzed in real time and doesn’t should be stored. This device would only require a smartphone, so it could be easy to put in and a cheap solution.
The idea is that first responders, akin to paramedics or emergency room nurses, have this app on their smartphone. When they see a patient with a medical emergency, they will use the app to find out inside seconds whether the patient could have suffered a stroke, and speed up treatment accordingly.
Our pilot study
We tested the tool on a small dataset using video recordings of 14 individuals who had suffered a stroke and 11 healthy controls.
We found that it had an accuracy of 82%, meaning it accurately identified a stroke 82% of the time. Our tool will not be intended to switch comprehensive clinical diagnostic tests for stroke, nevertheless it could help discover individuals who need treatment much earlier and support doctors.
Although these results are promising, we plan to proceed to optimize the model. We hope that accuracy will improve as we create a bigger dataset with records from more patients.
At this stage, the AI model has only been trained and developed on a small dataset, and the info lacks ethnic and demographic diversity. It will probably be vital to refine and test the app on people of various cultural and ethnic backgrounds.
In the long run, we plan to work with clinics, emergency departments and emergency services to conduct clinical trials. We have to test the effectiveness of this tool within the hands of actual users, akin to paramedics, to verify that it helps them take care of their patients.
