Predicting response to neoadjuvant therapy in breast cancer using longitudinal DCE-MRI deep learning integrated with
Lan Yan1, Xianming Huang2, Lan Liu1
1Department of Radiology, Jiangxi Cancer Hospital & Institute, Jiangxi Clinical Research Center for Cancer, The Second Affiliated Hospital of Nanchang Medical College, Nanchang, China.
Frontiers in Immunology
|May 14, 2026
Summary
A new multimodal model accurately predicts pathological complete response (pCR) in breast cancer patients receiving neoadjuvant therapy (NAT). This approach combines deep learning imaging features, inflammatory markers, and tumor-infiltrating lymphocytes for improved early treatment prediction.
Area of Science:
- Oncology
- Medical Imaging
- Artificial Intelligence
Background:
- Accurate prediction of pathological complete response (pCR) is crucial for optimizing neoadjuvant therapy (NAT) in breast cancer.
- Current prediction methods often lack accuracy and timeliness, hindering personalized treatment strategies.
Purpose of the Study:
- To develop and validate a multimodal fusion model for early and accurate prediction of pCR to NAT in breast cancer.
- To integrate deep learning (DL) features from dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI), peripheral blood inflammatory (PBI) indices, and tumor-infiltrating lymphocytes (TILs).
Main Methods:
- Retrospective analysis of 262 breast cancer patients undergoing NAT.
- Construction of DL models using pre-treatment and early-treatment DCE-MRI.
- Development of an immune-inflammation model using TILs and PBI indices.
- Integration of all modalities into a combined model for performance assessment.
Main Results:
- The combined model achieved a superior area under the receiver operating characteristic curve (AUC) of 0.90 and 95% specificity in the validation cohort.
- The combined model significantly outperformed single-modality models.
- Early-treatment DCE-MRI DL features showed critical predictive value (AUC=0.85), outperforming baseline DL features (AUC=0.75).
Conclusions:
- A multimodal fusion model integrating DL-based DCE-MRI features, dynamic inflammatory indicators, and baseline TILs significantly improves early pCR prediction in breast cancer NAT.
- This strategy offers a promising tool for personalized treatment planning in breast cancer patients receiving NAT.