Queensland artificial intelligence technology has joined forces with the state's world-leading medical research institute to tailor cancer patients' treatment and improve its effectiveness.
Subscribe now for unlimited access.
$0/
(min cost $0)
or signup to continue reading
The collaboration is an Australian first, according to the researchers involved, and will span at least two years after receiving $2.6 million in funding from the federal government last week.
Brisbane AI company Max Kelsen has partnered with the internationally renowned QIMR Berghofer Medical Research Institute, precision-analytics start-up genomiQa, genomics researcher BGI Australia and the Royal Brisbane and Women's Hospital.
Cancer patients' genetics will be analysed by AI in the hope of finding more complex patterns and therefore helping doctors to decide which treatment would be most effective for the individuals.
"The problem with genomic info is that it’s extremely complex," Max Kelsen CEO Nicholas Therkelsen-Terry said.
"Finding patterns is very different for humans; previously we have looked for clear markers ... but now we will use AI to look for much more complex interactions."
Mr Therkelsen-Terry said his company had worked on how to feed the complex genetic information into an AI algorithm for the past year and he was confident his technology could get results.
"We will be looking at new immunotherapy treatments, but the problem with it is the response rate, especially in small-cell lung cancer and melanoma.
"We are going to feed the genetic information of cancer patients who have and have not responded into AI algorithms and then hopefully get the right drug for them."
genomiQa CEO Colin Albert said the project aimed to save patients money — some treatments cost as high as $130,000 a year — and lead to government funding being spent more efficiently.
"In general, only 50 per cent of melanoma patients respond positively to immunotherapy and there is about a 30 per cent success rate for other cancers," he said.
"This research can lead to someone taking a $5000 test and that means the likelihood of treatment working could increase by 70-80 per cent.
"It could also mean government funds could be spent more efficiently and better spending will result in the government not wasting its budget on medications that might not work."
Mr Albert said the project was not just about science; it has also brought together computers and data analysis.
"The project is not led by a pharmaceutical company or a hospital; the lead is an AI company and that shows the shift in thinking from the medical community ... to see how to improve," he said.
"This is definitely an Australian first in terms of the combination of AI and the fact that we are doing this with new patients."
Mr Therkelsen-Terry said he hoped to investigate the potential commercialisation of the analysis to improve patient outcomes sooner.
"This is huge. In my opinion, AI will have an increased role in genomic analysis," he said.
"It's not just about which drug will respond best, but which combinations of drugs and what dosages will be most effective. The treatment can be fine-tuned by AI."
