Abstract
AIM: To develop and validate a radiomics-based nomogram using multimodal magnetic resonance imaging (MRI) features to predict HER-2 expression status in breast cancer. METHODS: A total of 320 breast cancer patients were retrospectively selected for this study, with 80 in the HER-2 positive group and 240 in the HER-2 negative group. Pre-treatment multimodal MRI scans, including dynamic contrast-enhanced MRI (DCE-MRI), diffusion-weighted imaging (DWI), and T2-weighted imaging, were used to extract radiomic features. Multivariate logistic regression was performed to identify independent predictors for HER-2 positivity. A radiomics-based nomogram was constructed and validated using both training and validation sets. The nomogram's performance was assessed using receiver operating characteristic (ROC) curve analysis and decision curve analysis (DCA). RESULTS: Multivariate analysis identified several independent predictors of HER-2 positivity, including type, edge, local skin thickening or depression, and axillary lymph node enlargement. The radiomics-based nomogram demonstrated excellent predictive accuracy with an area under the ROC curve (AUC) of 0.866 in the training set and 0.876 in the validation set. Calibration plots confirmed the model's good consistency, and DCA indicated that the nomogram provides significant clinical benefit across a range of threshold probabilities. In addition, the HER-2 positive group showed significantly higher tumor marker expression and immune cell infiltration, including elevated CD8+ T-cells, M1 macrophages, Tregs, and TAM (p<0.001). CONCLUSION: The radiomics-based nomogram developed in this study offers a promising non-invasive tool for predicting HER-2 expression status in breast cancer. By integrating clinical data and advanced MRI features, this model provides accurate predictions, improving personalized treatment strategies. Further validation in larger, multicenter studies is necessary to confirm its generalizability and clinical applicability.