Clinical Report: Combining Intratumoral and Peritumoral Radiomics with Deep Transfer Learning

Overview

This study evaluates a predictive model for HER2 status in breast cancer using a combination of clinical indicators, intratumoral and peritumoral radiomics, and deep transfer learning from multiparametric MRI. The model demonstrated high accuracy, achieving AUCs of 0.965 in the training cohort and 0.904 in the internal validation cohort.

Background

Breast cancer is the most prevalent malignancy among women and a leading cause of cancer-related mortality. Accurate determination of HER2 status is critical for effective treatment, as HER2-positive patients benefit significantly from targeted therapies. Current methods for assessing HER2 status are invasive and time-consuming, highlighting the need for non-invasive alternatives.

Data Highlights

Key Findings

• The combined model integrates clinical indicators with radiomics and deep learning features.
• High predictive performance was observed with AUCs exceeding 0.9 in validation cohorts.
• Feature selection involved intraclass correlation coefficient assessment and LASSO logistic regression.
• Grad-CAM and SHAP were utilized for model interpretability.
• Multiparametric MRI provided a non-invasive assessment method for HER2 status.

Clinical Implications

The model developed in this study offers a non-invasive method for predicting HER2 status.

Conclusion

The integration of radiomics and deep learning with clinical indicators presents a potential advancement in the preoperative assessment of HER2 status in breast cancer.

Related Resources & Content

1. European Radiology, 2024 -- Evaluation of Anti-HER2 Treatment Response for Tailoring Therapy in Early HER2-Positive Breast Cancer Utilizing an Innovative Deep Learning Radiomics Approach
2. Frontiers in Oncology, 2026 -- Intra- and peritumoral radiomics for predicting equivocal HER2 status of breast cancer on contrast-enhanced mammography
3. European Radiology, 2025 -- Radiomics Utilizing Deep Learning Fails to Enhance Prediction of Residual Cancer Burden Following Chemotherapy in the LIMA Breast MRI Study
4. Frontiers in Medicine, 2026 -- Multi-parametric MRI habitat radiomics with interpretable machine learning for early prediction of axillary lymph node metastasis in triple-negative breast cancer
5. HER2 Testing In Breast Cancer - 2023 Guideline Update - CAP
6. Trastuzumab Deruxtecan after Endocrine Therapy in Metastatic Breast Cancer.
7. Combining intratumoral and peritumoral multimodality magnetic resonance imaging (MRI) to predict the expression level of human epidermal growth factor receptor-2 (HER-2) in breast cancer - PubMed
8. HER2 Testing In Breast Cancer - 2023 Guideline Update - CAP
9. Trastuzumab Deruxtecan after Endocrine Therapy in Metastatic Breast Cancer.
10. Combining intratumoral and peritumoral multimodality magnetic resonance imaging (MRI) to predict the expression level of human epidermal growth factor receptor-2 (HER-2) in breast cancer - PubMed