Western Health is set to try out a new AI-powered tool that factors in patients’ body compositions to calculate their precise chemotherapy dosing
Researchers from the University of Melbourne and Western Health recently developed a digital tool that analyses CT scans of cancer patients using image recognition and AI and tailors their chemotherapy doses based on their body compositions.
Findings from their study, published in 2023, showed that body compositions, or the percentages of fat, bone and muscle, determined how a chemotherapy drug was metabolised and stored in patients’ bodies.
“Our algorithm was able to produce accurate chemotherapy dosing for 84% of those patients, which is a significant improvement over current methods of dosing,” noted Justin Yeung, study lead and a surgery professor at the University of Melbourne. Chemotherapy doses are typically calculated using patients’ body surface area based on height and weight.
“It doesn’t make logical sense for two patients with significantly different fat and muscle ratios to be given the same chemotherapy dose just because they have similar body surface areas,” stressed Prof Yeung, who is also a consultant colorectal surgeon at Western Health.
“For example, a sumo wrestler and a body builder who have comparable body surface areas would theoretically be given the same chemotherapy doses; however, as their body compositions are vastly different, they would likely develop different degrees of toxicities,” he further explained.
The team has received nearly A$500,000 in grant funding from Australia’s Economic Accelerator program for their AI development and trial.
WHY IT MATTERS
The research team, which has formed a startup called PredicTx Health to turn their chemotherapy dosing solution into a commercial product, developed their AI algorithm using CT scan data from a cohort of over 1,000 colorectal cancer patients at Western Health.
They initially focused on colorectal cancer, given that 60% of patients who undergo chemotherapy for this condition were either overdosed or underdosed. Overdosing can cause a range of side effects, including immunosuppression, heart attacks and chest infections.
Due to these side effects, they noted, patients tend to stop their treatment early, thereby significantly reducing their chance of survival.
THE LARGER TREND
Recently, St Vincent’s Hospital Melbourne started utilising an AI-driven analyser to help verify the identity and concentration of intravenous drugs at its compounding pharmacy. It said the verification solution adds an extra layer to prevent medication error at the point of compounding.
Meanwhile, many health services across Victoria have digitised their cancer information management systems. Peninsula Health, for example, adopted one of the widely recognised solutions, Charm Evolution by Magentus, to automate workflows. Slade Health, a compounding pharmacy under cancer care provider Icon Group, implemented the same solution to digitise its ordering system.
