ObjectiveTo investigate the value of radiomics nomogram based on DCE-MRI in predicting PD-L1 expression in breast invasive ductal carcinoma of breast.

MethodsA retrospective analysis was conducted in 91 patients with invasive ductal carcinoma who underwent preoperative breast MRI examination and confirmed by pathology at The First Affiliated Hospital of Bengbu Medical College.The patients were randomly divided into a training set (n=54) and a test set (n=37) at the ratio of 6∶4.The second stage of DCE-MRI was selected to delineate the 3D region of interest (ROI) layer by layer at all levels of lesion.The optimal radiomics features were screened through minimum absolute contraction and regression with the least absolute shrinkage and selection operator, the radiomics score (Rad-score) was obtained by support vector machine to construct the radiomics model, independent risk factors of clinical and imaging features were screened through univariate and multivariate logistic regression to construct a clinical model, and Rad-score combined with risk factors of clinical and imaging features was used to construct a radiomics nomogram model finally.The prediction efficacy of each model was analyzed by the receiver operating characteristic curve, and the corresponding area under the curve (AUC) was calculated to obtain the optimal model.The predictive efficacy and clinical utility of the model were evaluated using the calibration curve and the decision curve.

ResultsThe radiomics nomogram model had the best diagnostic performance.The AUC, sensitivity, and specificity of the nomogram model were 0.938, 81.8%, 95.2% for the training set and 0.904, 65.2%, 100.0% for the test set, respectively.

ConclusionsThe radiomics nomogram model based on DCE-MRI is highly valuable in predicting the PD-L1 expression in invasive ductal carcinoma of breast, and is helpful to evaluate the treatment effect and prognosis of patients.