Deep learning prediction of pathological complete response in breast cancer using Mamba architecture
A deep learning model helps predict which breast cancer patients will respond completely to chemotherapy before treatment starts, across multiple hospitals.
A novel Mamba-architecture deep learning model predicts pathological complete response to neoadjuvant chemotherapy in breast cancer from routine needle biopsy images with AUROC up to 0.923 in training and 0.761-0.809 across four independent external hospital cohorts. This multi-hospital validation across 1646 patients suggests clinical readiness for decision-support integration to help tailor NAC regimens.
What the study was
- Study design
- Multi-center retrospective validation study
- Population
- Breast cancer patients receiving neoadjuvant chemotherapy (NAC), 5 Chinese tertiary hospitals
- Sample size
- 1646 patients (1023 training/validation, 623 external test across 4 hospitals)
- Category
- Diagnostics
- Maturity
- Validated
- Journal
- NPJ Digital Medicine
Why it surfaced
Large multi-center validation (n=1646) of a novel Mamba-architecture AI model for predicting breast cancer pCR after NAC. Robust external validation across 4 independent hospitals is a strong signal of clinical applicability. pCR prediction has direct treatment decision implications (surgery timing, regimen continuation/change).
A plain-language summary of published research — not medical advice. Talk to a clinician about your own care.