What happened
The University of Adelaide reported that EndoFusion, a framework developed through the IMAGENDO study, detected two imaging signs of advanced endometriosis using pelvic MRI and transvaginal ultrasound data. Researchers said the model reached a correct diagnosis 83% of the time and processed scans in milliseconds.
Why it matters
Endometriosis diagnosis can be slow and invasive. Tools that make imaging more consistent could matter, but early model performance is not the same as clinical readiness, and broader validation is needed across sites, operators and disease presentations.
What to watch
External validation, performance in adolescents and earlier disease, false positives and false negatives, workflow integration, regulatory pathway and whether use reduces time to diagnosis.
