Abstract:Objective: To study the diagnostic efficacy of the nomogram model constructed by contrast-enhanced ultrasound combined with risk of ovarian malignancy algorithm (ROMA) index in epithelial ovarian cancer (EOC). Methods: A retrospective analysis was implemented on 214 patients who were initially diagnosed with ovarian tumors and received surgical treatment in the gynecology department of Gansu Provincial Maternity and Child-care Hospital from April 2019 to April 2022. About 130 patients recruited from April 2019 to December 2020 were used as the modeling set, and 84 patients recruited from January 2021 to April 2022 were used as the validation set. Based on the results of pathological diagnosis, the modeling set was divided into benign tumor group (n=85) and EOC group (n=45). The ROMA index and quantitative parameters of contrast-enhanced ultrasound were contrasted between the two groups, and the differential factors were selected by logistic regression to construct a nomogram diagnostic model. Results: Compared with the benign tumor group, the positive ratio of CA125, HE4, and ROMA, and the value of PI, V1 and TIC-AUC in the EOC group were significantly increased (P<0.05), while the value of IT and TTP were significantly decreased (P<0.05). Multivariate logistic regression results showed that ROMA, IT, TTP, PI, and TIC-AUC were the different factors between EOC patients and benign tumor patients (P<0.05). The areas under the ROC curves of the modeling set and validation set were 0.919 (95%CI: 0.873-0.965, P<0.001) and 0.835 (95%CI: 0.737-0.933, P<0.001), respectively. When the diagnostic probability was taken as the optimal cut-off value of 0.244, the sensitivity, specificity and accuracy of the nomogram diagnostic model were 86.67%, 80.00% and 82.31%, respectively. Conclusion: The nomogram model constructed by the quantitative parameters of contrast-enhanced ultrasound combined with the ROMA index has better discrimination and calibration, it also has higher diagnostic value for EOC patients.