AI system aims for earlier diagnosis of endometriosis

IMAGENDO uses machine learning to digitally combine the diagnostic capabilities of pelvic scans and MRI to identify endometriosis lesions. The Australian and U.K. developers are now seeking to optimize and validate the algorithm.

The IMAGENDO system is being developed by scientists at the Robinson Research Institute at the University of Adelaide, in partnership with colleagues at the University of Surrey in the U.K.

IMAGENDO uses machine learning to digitally combine the diagnostic capabilities of pelvic ultrasound scans and MRI to identify endometriosis lesions. The researchers hope AI can shorten the diagnostic journey of millions of women suffering from endometriosis by reducing avoidable hospitalizations and repetitive surgery and improving treatment pathways.

The lengthy diagnosis process for endometriosis -- it can take six years on average -- can lead to anxiety, depression, and fatigue. Without a reliable noninvasive test, keyhole surgery may be necessary to visualize endometrial deposits inside the abdomen, ideally verified by microscopic examination of the tissue.

The scientists are using AI to analyze data from ultrasound and MRI scans to try to provide fast, noninvasive diagnosis for people with endometriosis, a painful and common reproductive disease where sensitive tissue grows outside of the uterus.

"While the legitimate concerns about the use of AI have dominated the headlines, here is an example of how this technology can improve the lives of millions of people who suffer from endometriosis and severe pelvic pain,” Prof. Gustavo Carneiro, professor of AI and machine learning at the University of Surrey and one of the chief investigators of IMAGENDO, said in a statement.

The IMAGENDO system has been shortlisted for the prestigious ANSTO Eureka Prize for Innovative Use of Technology. Eureka Prizes are awarded by the Australian Museum every year to recognize individuals and organizations who have contributed to science in Australia.