MRI Radiomics with Extra Trees Classifier Accurately Distinguishes Endometriomas from Hemorrhagic Cysts

An MRI radiomics-based Extra Trees classifier differentiated ovarian endometriomas from hemorrhagic cysts with an AUC of 0.99 (95% CI 0.97–1.00), accuracy 96%, sensitivity 96%, and specificity 95% in 160 women. External validation is needed.

• Five radiomic features from T2-weighted MRI were selected as most discriminative.
• Ensemble classifiers outperformed single models (AUCs 0.97–0.99 vs 0.87–0.96).
• Single-center design without external validation limits generalizability.

Automated summary

RadPigeon summaries are original and for information only. They are not clinical advice.