Kurhan - stock.adobe.com
In late 2019, a team of researchers led by Professor John Fraser from University of Queensland’s critical care research group planned to conduct a small study to ascertain the next influenza outbreak in the Asia-Pacific region.
It didn’t take too long before the plan was scuttled in January 2020, when news of the Covid-19 coronavirus outbreak emerged. The researchers then decided to pivot towards a new study to determine the best treatments for the novel disease.
“Even before the virus was named, we were hearing stories from our friends and colleagues that the virus didn’t seem like flu,” said Fraser at the inaugural IBM Cloud Forum 2020. “Lungs were reacting differently; patients were leaning towards kidney failure and were waking up with bizarre delirium.”
Experienced clinicians and doctors across the region who did not know what was happening to their patients and how to ventilate them started asking Fraser’s team for advice. With no clue on how the disease would pan out, the team realised they had to do something to save lives before a vaccine was found.
“At that point, we realised that there were weapons we could use,” he said. “There were about maybe 20,000 intensive care patients out there already – if only we could somehow bring the data on those patients together to help us.”
Fraser likened the effort to putting together a jigsaw puzzle with 20,000 pieces of human data spread across the globe. Unveiling the full picture of the situation, he said, would require a dashboard with machine learning capabilities.
“Dashboards are the easiest and fastest way to update the data on a daily basis, so that we can create information that could guide clinicians to create their own decision support,” said Fraser. “And through artificial intelligence [AI] and machine learning, and as we feed the machine more and more data, the strength of this dashboard would become much greater.”
Read more about AI in Australia and New Zealand
- Using Amazon’s AI smarts, National Australia Bank has developed a unique brand voice that will greet customers who dial up its contact centre.
- Air New Zealand has teamed up with Auckland Airport to test the use of computer vision to turn around its fleet faster.
- Retailers in the Asia-Pacific region believe AI is key to their success, but lack the skills and infrastructure needed to harness the technology.
- A smart cricket ball powered by IoT and the machine learning smarts of AWS is helping amateur players improve their game.
Fraser’s team worked with Sally Shrapnel, a senior lecturer from the school of mathematics and physics at the same university to develop the dashboard, as well as IBM, which supplied the AI smarts in the form of IBM Watson, a suite of AI services, applications and tooling.
“The IBM team worked incredibly well with us,” said Fraser. “Rather than develop the technology in isolation, they spent countless hours with doctors in about 25 countries to get their feedback.”
With the dashboard, clinicians in some 300 intensive care units participating in the study will be able to quickly record and share key data points, including intensive care duration and survival rates, along with clinical insights on a range of pulmonary, cardiological, neurological and renal treatments.
“Just because we can’t fly doesn’t mean data can’t fly,” he said. “We are taking this data globally and creating not doctors without borders, but technology without borders.”
“Watson may not able to cure the pandemic, but the creation of networks, the collaboration across divides, whether that’s technology or medicine, and the combination of medical opinion leaders and industry will allow us to return to the near-normal life that we so desperately desire.”