Skip to main content
Medtech biomedical engineering students: Shu Huang, Aeryne Lee, Michelle Yi Fan Lu, Feixue Ma, Joshua Riley, James Spinks and Christine Trinh.
News_

Can computer vision and AI predict strokes and prevent disability?

22 November 2017
Rapid stroke prediction device wins MedTech Innovation Competition

More than 80% of strokes are preventable. University of Sydney biomedical engineering students have proposed a rapid stroke prediction system using computer vision and AI, and accessible to any hospital with a CT imaging machine. 

Stroke kills more women than breast cancer and more men than prostate cancer. According to the Stroke Foundation,[1] one Australian will have a stroke every nine minutes costing the economy around $5 billion a year. Some 65% of stroke survivors will suffer a disability which impedes their ability to carry out daily living activities unassisted.

More than 80% of strokes can be prevented and yet NSW hospitals do not see the benefits of current methods of predicting strokes due to their cost and poor response times. Working with Associate Professor Noel Young at Westmead Clinical School, a team of University of Sydney biomedical engineering students have designed a rapid stroke prediction system using computer vision and AI to address this pressing need. It was the winning proposal at this year’s MedTech Innovation Competition.

Hosted by the University of Sydney's Institute of Biomedical Engineering and Technology, the annual MedTech Innovation Competition pairs fourth-year biomedical engineering students with industry-based supporters such as Westmead Hospital, Biointellect, CSIRO, Fledge and Corin Orthopaedics. Working under the guidance of Westmead clinicians, 20 student teams developed innovative solutions which included a non-invasive intracranial pressure monitor, a fall prevention gyro, osteoelastic bone blade system and dementia screening.

“We learned of the immediate need for a stroke predictive system that can provide accurate results within six hours of onset,” says Joshua Riley, one of the winning team members.

The team’s ViSP System uses computer vision and analysis combined with 4D CT perfusion scans and deep machine learning to predict the chances of stroke. By analysing a variety of factors including tissue density, blood volumetric flow and velocity analysis, the system will generate a risk factor for ischaemic and haemorrhagic strokes which, when packaged into a detailed report for clinicians, may help determine the next steps in treatment.  

“The competition gave us a chance to put all the skills we’ve learnt over the past three years into action and really showed us what a career in biomedical engineering is all about,” says Christine Trinh, Engineering Honours (Biomedical) / Commerce student at the University of Sydney.

“All the projects undertaken for this competition have great potential to help millions of people, and with that, I’m thankful to have had this exposure to potentially life-changing technology.”

 

MedTech Innovation Competition winning team members as pictured: Shu Huang, Aeryne Lee, Michelle Yi Fan Lu, Feixue Ma, Joshua Riley, James Spinks and Christine Trinh.

[1] https://strokefoundation.org.au/About-Stroke/Facts-and-figures-about-stroke