Adelaide University researchers develop AI tool for endometriosis diagnosis

Researchers at Adelaide University have developed an artificial intelligence tool called EndoFusion designed to accelerate the diagnosis of endometriosis. The tool, part of the IMAGENDO collaborative study, can identify two major indicators of advanced endometriosis in pelvic scans in approximately 18 milliseconds.

Currently, medical professionals often rely on either MRI or ultrasound imaging, which may each excel at detecting different markers of the condition. This can lead to diagnostic disadvantages depending on which scan a patient receives. EndoFusion aims to address this by combining data from both imaging types, allowing for more effective detection through a single scan.

The AI framework was trained using datasets comprising over 9,000 female pelvic MRI scans and more than 800 transvaginal ultrasound scans. In testing, the tool achieved an 83% correct diagnosis rate, which researchers noted was more accurate than competing models. The technology is currently in the early stages of development.