Detecting strokes in real time with AI and a smartphone

The revolutionary technology spots real-time changes in facial symmetry

Spotted: Every year, 15 million people experience the life-altering effects of a stroke, which occurs when blood flow to the brain is interrupted, depriving brain tissue of essential oxygen and nutrients. Even a brief delay in treatment can cause permanent damage, making early detection crucial.

Aiming to revolutionise early stroke detection, a team of biomedical engineers at the Royal Melbourne Institute of Technology (RMIT) has developed cutting-edge AI software capable of identifying stroke symptoms using a smartphone.

A key sign of a stroke is unilateral facial muscle weakness, causing one side of the face to behave differently from the other. To address this, the RMIT team, led by Guilherme Camargo de Oliveira and supervised by team leader Professor Dinesh Kumar, developed AI-driven technology that monitors real-time changes in facial asymmetry. By analysing a short video taken on a smartphone, the software assesses facial symmetry and specific muscle movements, helping to quickly determine if immediate treatment is needed.

When talking to Springwise, Professor Dinesh Kumar explained that the team recognised the difficulty of early stroke symptom detection. Their smartphone solution shows promising potential to positively impact the healthcare system, save lives, and reduce chances of disability.

The next steps for the team focus on clinical trials, in a bid to secure Therapeutics Goods Administration (TGA) approval in Australia and enable the smartphone tool to start making a difference.

Written By: Georgia King