A modified ovarian-adnexal reporting and data system ultrasound for differentiating ovarian carcinoma: integration of the risk of malignancy algorithm into the American College of Radiology's O-RADS US framework.

IntroductionMost ovarian cancer cases are diagnosed at advanced stages, primarily due to the limited efficacy of current classification systems that rely on imaging modalities with variable accuracy. Transvaginal ultrasound as an imaging modality is commonly achieved using the detection of morphological features, which has some limitations mainly related to sonographer experiences and subjective nature of the assessment. To address this limitation, we developed a novel scoring system by incorporating the risk of malignancy algorithm (ROMA) into the ovarian-adnexal reporting and data system ultrasound lexicon (O-RADS US) framework, aiming to refine ultrasound-based risk stratification.MethodsThis retrospective study enrolled 151 patients (mean age, 51.6 years ± 15.3), with 159 ovarian-adnexal masses (median, 8.2 cm; range, 2.1-24.9cm) between January 2023 and May 2026. The masses were assessed according to O-RADS US criteria, while ROMA values were calculated preoperatively. Then, the modified O-RADS US system reclassified lesions by integrating ROMA-derived malignancy probabilities into O-RADS US categories.ResultsThe modified O-RADS US system demonstrated superior diagnostic efficiency for ovarian cancer detection, achieving an AUC of 0.922 (95% CI: 0.878-0.966), significantly outperforming both the original O-RADS US system (AUC = 0.815, 95% CI: 0.740-0.890; p < 0.05) and ROMA alone across premenopausal (AUC = 0.861, 95% CI: 0.812-0.910; p 0.05), and combined cohorts (AUC = 0.894, 95% CI: 0.865-0.923; p > 0.05). A marked reduction in O-RADS 4 lesions and an increase in O-RADS 2/3 lesions, while O-RADS 5 lesions remained stable (23.3% vs. 23.9%), and the malignancy rate for Categories 4 and 5 lesions increased from 44.0% with the original system to 58.1% with the modified system.ConclusionThe incorporation of ROMA into the ACR O-RADS US framework significantly enhances diagnostic performance for ovarian carcinoma, and improves risk stratification accuracy, reducing unnecessary biopsies for low-to-moderate risk lesions.