Objective To investigate the value of nomogram constructed based on dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) combined with clinical features in predicting the expression status of vasculogenic mimicry (VM) in patients with triple-negative breast cancer (TNBC).
Methods Ninety-four cases of breast cancer diagnosed by DCE-MRI before operation and confirmed by pathology were analyzed retrospectively, and randomly assigned into a training set (n=65) and a test set (n=29) at a ratio of 7∶3.DCE-MRI phase 2 was selected to delineate the maximum lesion level.The optimal features were selected by f-calssif function, least absolute shrinkage and selection operator regression, and the DCE-MRI radiomics model was constructed by support vector machine (SVM).The independent clinical predictors were screened by single-multiple logistic regression to build the clinical model, and Rad-score of DCE-MRI model combined with independent clinical predictors was selected to build the nomogram model.
Results In the training set, there were statistically significant differences in the presence or absence of axillary lymph node (ALN) metastasis, MRI maximum diameter difference, and tumor margin between patients with positive and negative VM expression (P < 0.05 to P < 0.01).In the nomogram model, the AUC, sensitivity, specificity, and accuracy of the training set were 0.880, 82.4%, 89.6%, and 98.5%, respectively, and which of the test set were 0.869, 87.5%, 81.0%, and 82.8%, respectively.The nomogram model had good consistency with the ideal model in judging the training set and testing ALN metastasis results (P < 0.05).
Conclusions Nomogram based on DCE-MRI can be used as an accurate and non-invasive method to predict VM expression levels in preoperative TNBC patients.
DownLoad: