Background Preoperative differentiation between benign and borderline epithelial ovarian tumors (EOTs) is challenging and can significantly impact clinical decision making. The purpose was to investigate whether radiomics based on T2-weighted MRI can discriminate between benign and borderline EOTs preoperatively. Methods A total of 417 patients (309, 78, and 30 samples in the training and internal and external validation sets) with pathologically proven benign and borderline EOTs were included in this multicenter study. In total, 1130 radiomics features were extracted from manually delineated tumor volumes of interest on images. The following three different models were constructed and evaluated: radiomics features only (radiomics model); clinical and radiological characteristics only (clinic-radiological model); and a combination of them all (combined model). The diagnostic performances of models were assessed using receiver operating characteristic (ROC) analysis, and area under the ROC curves (AUCs) were compared using the DeLong test. Results The best machine learning algorithm to distinguish borderline from benign EOTs was the logistic regression. The combined model achieved the best performance in discriminating between benign and borderline EOTs, with an AUC of 0.86 ± 0.07. The radiomics model showed a moderate AUC of 0.82 ± 0.07, outperforming the clinic-radiological model (AUC of 0.79 ± 0.06). In the external validation set, the combined model performed significantly better than the clinic-radiological model (AUCs of 0.86 vs. 0.63, p = 0.021 [DeLong test]). Conclusions Radiomics, based on T2-weighted MRI, can provide critical diagnostic information for discriminating between benign and borderline EOTs, thus having the potential to aid personalized treatment options.

PurposeMachine learning (ML) can extract high-throughput features of images to predict disease. This study aimed to develop nomogram of multi-parametric MRI (mpMRI) ML model to predict the risk of breast cancer.MethodsThe mpMRI included non-enhanced and enhanced T1-weighted imaging (T1WI), T2-weighted imaging (T2WI), apparent diffusion coefficient (ADC), Ktrans, Kep, Ve, and Vp. Regions of interest were annotated in an enhanced T1WI map and mapped to other maps in every slice. 1,132 features and top-10 principal components were extracted from every parameter map. Single-parametric and multi-parametric ML models were constructed via 10 rounds of five-fold cross-validation. The model with the highest area under the curve (AUC) was considered as the optimal model and validated by calibration curve and decision curve. Nomogram was built with the optimal ML model and patients’ characteristics.ResultsThis study involved 144 malignant lesions and 66 benign lesions. The average age of patients with benign and malignant lesions was 42.5 years old and 50.8 years old, respectively, which were statistically different. The sixth and fourth principal components of Ktrans had more importance than others. The AUCs of Ktrans, Kep, Ve and Vp, non-enhanced T1WI, enhanced T1WI, T2WI, and ADC models were 0.86, 0.81, 0.81, 0.83, 0.79, 0.81, 0.84, and 0.83 respectively. The model with an AUC of 0.90 was considered as the optimal model which was validated by calibration curve and decision curve. Nomogram for the prediction of breast cancer was built with the optimal ML models and patient age.ConclusionNomogram could improve the ability of breast cancer prediction preoperatively.