Clinical Report: Interpretable Machine Learning-Enhanced Multi-Parametric MRI Habitat Radiomics for Early Detection of Axillary Lymph Node Metastasis in Triple-Negative Breast Cancer

Overview

This study developed and validated a multiparametric MRI-based habitat radiomics model for predicting axillary lymph node metastasis (ALNM) in triple-negative breast cancer (TNBC). The combined model demonstrated superior predictive performance, highlighting the importance of intratumoral heterogeneity in non-invasive nodal risk stratification.

Background

Accurate preoperative assessment of ALNM in TNBC is crucial for treatment planning, as nodal involvement is linked to poor survival and recurrence risk. Current invasive methods for axillary staging underscore the need for non-invasive predictive tools. This study addresses the clinical gap by utilizing advanced imaging techniques to enhance predictive accuracy.

Data Highlights

Key Findings

• The combined model achieved the highest AUC of 0.81 in the test set for predicting ALNM.
• The habitat radiomics model outperformed both conventional radiomics and clinical models.
• SHAP analysis identified ALN length as the most significant predictor of ALNM.
• Intratumoral heterogeneity characterization is essential for non-invasive nodal risk stratification.
• Multiparametric MRI provides valuable insights into tumor biology relevant for ALNM prediction.

Clinical Implications

The findings suggest that integrating habitat radiomics into clinical practice can enhance preoperative assessments of ALNM in TNBC patients. This non-invasive approach may reduce the need for invasive staging procedures, aligning with current trends toward de-escalation of axillary surgery.

Conclusion

The study demonstrates that an interpretable habitat radiomics model significantly improves the prediction of ALNM in TNBC, offering a promising tool for non-invasive risk assessment. This advancement could facilitate better treatment planning and patient outcomes.

Related Resources & Content

1. Frontiers in Oncology, 2026 -- MRI-based habitat radiomics for preoperative prediction of axillary pathological complete response in breast cancer after neoadjuvant therapy: a multicenter study
2. Int. Journal of Computer Assisted Radiology and Surgery (Springer), 2026 -- Estimation of histopathological types from breast MRI findings using a large language model
3. conexiant -- AI Model May Help Decipher Malignant, Benign Breast Lesions on MRI
4. European Radiology -- Radiomics Utilizing Deep Learning Fails to Enhance Prediction of Residual Cancer Burden Following Chemotherapy in the LIMA Breast MRI Study
5. Sentinel Lymph Node Biopsy in Early-Stage Breast Cancer: ASCO Guideline Update | Journal of Clinical Oncology
6. ACR Appropriateness Criteria® Imaging of the Axilla - PubMed
7. Intratumoral and peritumoral habitat imaging based on multiparametric MRI to predict HER2-negative breast cancer subtypes: a multicenter study | BMC Medical Imaging | Springer Nature Link
8. Sentinel Lymph Node Biopsy in Early-Stage Breast Cancer: ASCO Guideline Update | Journal of Clinical Oncology
9. ACR Appropriateness Criteria® Imaging of the Axilla - PubMed
10. Intratumoral and peritumoral habitat imaging based on multiparametric MRI to predict HER2-negative breast cancer subtypes: a multicenter study | BMC Medical Imaging | Springer Nature Link