A deep learning system for non-invasive breast cancer diagnosis with multimodal data
Combining ultrasound and mammography through AI improves breast cancer detection accuracy without needing tissue biopsies, tested across multiple hospitals.
A deep learning system published in Nature Biomedical Engineering integrates multimodal imaging data (ultrasound and mammography) for non-invasive breast cancer diagnosis, with multicenter validation across several Chinese cancer hospitals. The system was developed by ShanghaiTech University and collaborating institutions and represents a potentially scalable tool for improving breast cancer detection accuracy without tissue biopsy.
What the study was
- Study design
- Multicenter diagnostic validation study (multimodal imaging DL system)
- Population
- Patients undergoing breast imaging at multiple Chinese cancer centers and hospitals
- Category
- Early Detection
- Maturity
- Validated
- Journal
- Nat Biomed Eng
Why it surfaced
Nat Biomed Eng multimodal DL system for non-invasive breast cancer diagnosis with multicenter validation. Abstract was truncated at author affiliations during efetch — confidence set to medium; journal and institutional context confirm high-impact imaging AI study.
A plain-language summary of published research — not medical advice. Talk to a clinician about your own care.