Using AI to develop cancer treatments

This is the result of four years of research by a research team led by Associate Professor Alex Gavryushkin at the Centre for Mathematical Biology at the University of Canterbury.

In the study, experts developed algorithms that analyze biological data related to complex genetic diseases - such as cancer and gout - to help develop treatment regimens based on genetic data.

“The traditional medical approach often looks at many patients with a particular condition to choose the right treatment,” said Associate Professor Gavryushkin. “However, for many diseases, including cancer, the condition of each patient is very different even though they have similar symptoms. Therefore, applying the same drugs and therapies will not be effective.” “The answer to that is genomics,” he stressed.

Associate Professor Gavryushkin said his team developed and “trained” their algorithms on genomic and clinical data to match each patient’s condition with known knowledge and methods in biology and medicine. They then conducted clinical trials to come up with useful recommendations, such as combining different drugs in treatment.

He believes that this technology will allow medical professionals to develop more effective cancer treatment regimens, curb the development of treatment-resistant cells, and make medical care more convenient. Therefore, the team's research focuses mainly on prioritizing and planning treatments in a way that minimizes the risk of developing treatment-resistant genotypes. The worst risk that occurs after treatment is that the population of treatment-resistant cells will grow back.

Associate Professor Gavryushkin said the AI ​​approach could serve as a reliable medical assistant for oncologists, and he expected it to be used globally, especially in places where doctors may not have extensive genetic training or the time to study all the previous genetic literature.

The next phase of the research will focus on testing the new technology in interactions with patients, oncologists and epidemiologists. In addition, Associate Professor Gavryushkin’s team is also applying the algorithms to treat gout, another complex genetic disease. Both research projects involve collaboration with researchers and medical experts from New Zealand, Switzerland, Spain and the United States.