The IMAGENDO study aims to use machine learning and an AI algorithm to make the process faster and more painless for patients.
A University of Adelaide study using AI to develop a non-surgical method of diagnosing endometriosis has been shortlisted for an Australian Museum Eureka Prize.
The IMAGENDO project, which is trialling machine learning on specialist ultrasounds and MRI scans to diagnose endometriosis earlier and with greater accuracy, was one of three finalists selected for the ANSTO Prize for Innovative Use of Technology.
According to program manager Dr Jodie Avery, the algorithm is still in the early stages of development, with the team working on having it identify a number of different signs of endometriosis before it is validated in several patient cohorts nationwide.
However, it has the potential to significantly reduce the time it takes to receive an endometriosis diagnosis and allows clinicians to introduce preventative interventions much earlier, with a national and international rollout anticipated within the next three to four years, Dr Avery said.
“At the moment, there’s a 6.5-year delay for women to get a diagnosis with endometriosis. If we can reduce that diagnostic delay, [then doctors] can make some important clinical decisions in a suitable time period, rather than waiting for such a long time,” Dr Avery told The Medical Republic.
“On the preventative side of things, if we know the stage of the endo non-invasively rather than doing surgery, we can do things like egg freezing, or if [a patient] is having repeated attacks, we can put some treatments in that might stop [the disease] from progressing as far over time.”
The team is currently considering a number of ways the technology could be rolled out in clinical practice, including integrating it with ultrasound machines to allow for real-time diagnoses or establishing an online, cloud-based platform where patients could upload their scans and receive a diagnosis through an integrated health record system.
However, working out how this implementation would be funded is a major challenge, while the high costs of ultrasounds and small number of sonographers trained to provide them could potentially limit uptake among patients, Dr Avery said.
“These scans can be quite expensive and they take up a lot of time, so we have to make sure we implement the training programs that are needed [to get] enough sonographers able to do these scans.”
