Intelligent machines to support, not replace, doctors
Artificial intelligence and medical machine learning are constantly evolving, however researchers at the University of Adelaide don’t expect these technologies to replace medical doctors any time soon.
Medical machine learning involves the development of new algorithms and models, which can then interpret medical data and improve clinical diagnosis and prognosis.
Currently, our researchers apply medical machine learning to the healthcare areas of cardiology, colorectal cancer screening, obstetrics and gynaecology, arthroscopy, public health, stroke, vascular dementia, and breast cancer screening.
Professor Anton van den Hengel, Director of the Australian Institute for Machine Learning, says that while the institute is continuously working on exciting new projects that will benefit medical practice, he doesn’t anticipate AI replacing doctors altogether in the future.
“I really don’t think any of this is replacing doctors, it will only help them make better decisions and help them spend more time focusing on what they are really good at,” said Professor van den Hengel.
“That is interacting with patients and figuring out patient priorities and how else they can help.”
Among the success stories coming out of the Institute is LBT Innovations, an Adelaide-based company that is now producing an ‘entirely new class of medical device’ that is being sold in the US. The device enables sophisticated AI to be applied to data captured elsewhere, supporting pathology and delivering better patient outcomes.
Other successful applications of medical machine learning includes developing new ways to interpret chest X-rays, retinal images and mammograms. Breast cancer screening in particular is set to experience an increase in accurate diagnosis by up to 10% thanks to a new AI system development from a research project led by Dr Gabriel Maicas and Professor Gustavo Carneiro.
Their system removes the complexities of previous systems by scanning the entire breast and not just ‘suspicious areas’, a development that will be particularly beneficial to younger patients with a possible hereditary link to breast cancer.
Professor Gustavo Carneiro
Australian Institute for Machine Learning
Dr Gabriel Maicas
ARC Research Fellow