Related Experiment Video
Updated: Aug 5, 2026
Competing-Risk Nomogram for Predicting Cancer-Specific Survival in Multiple Primary Colorectal Cancer Patients after Surgery
Published on: September 27, 2024
Development and validation of a multimodality radiomics-based nomogram for predicting HER2 expression status in
Xianwei Yang1,2, Xiaoling Liu1, Shenghong Wan1
1Department of Radiology, Affiliated Hospital of Guizhou Medical University, Guiyang, China.
Background:
This study aimed to develop and validate a nomogram integrating multimodal imaging and clinical features for the preoperative prediction of human epidermal growth factor receptor 2 (HER2) expression status in patients with invasive breast cancer.
Methods:
A total of 204 patients with pathologically confirmed breast cancer who underwent preoperative dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) and mammography (MG) were retrospectively enrolled. Patients were randomly divided into training and test sets in a 7:3 ratio. Pearson correlation coefficient (PCC) combined with recursive feature elimination (RFE) was used to select optimal radiomics features. Three binary classification tasks were constructed (Model_1: HER2-Positive vs. HER2-Zero; Model_2: HER2-Positive vs. HER2-Low; Model_3: HER2-Low vs. HER2-Zero), each containing three sub-models based on MG, DCE-MRI, and their combination. The optimal radiomics model for each task was selected and integrated with independent clinicopathological factors to establish a clinical-radiomics nomogram.
Results:
The combined radiomics model achieved the best performance for Model_1 and Model_2, with area under the curve (AUC) values of 0.742 (95% confidence interval [CI]: 0.624-0.860) and 0.823 (95% CI: 0.749-0.897) in the training set, and 0.718 (95% CI: 0.594-0.842) and 0.778 (95% CI: 0.696-0.861) in the validation set, respectively. Corresponding nomograms were subsequently constructed incorporating independent clinical predictors. For Model_3, the single DCE-MRI model demonstrated optimal performance (training AUC = 0.831, 95% CI: 0.747-0.916; validation AUC = 0.745, 95% CI: 0.640-0.850). However, no nomogram was developed due to the absence of significant clinicopathological features (p > 0.05).
Conclusion:
The multimodal radiomics-based nomogram, integrating imaging features and clinical factors, provides a promising non-invasive quantitative tool for preoperative evaluation of HER2 expression status in breast cancer, offering significant potential for clinical translation.