BAMT-Net: a boundary-aware multi-task framework for 3D breast ultrasound lesion segmentation and classification
Shuang He, Dan Ji, Mingwei Ma, Xiang Pan, Juxiang Xu, Feng Liu · 2026
Read BAMT-Net: a boundary-aware multi-task framework for 3D breast ultrasound lesion segmentation and classification by Shuang He, Dan Ji, Mingwei Ma, Xiang Pan, Juxiang Xu, Feng Liu (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/pmc-13269559.txt.
Description
Three-dimensional breast ultrasound gives doctors a full volume to search, but reading it by hand is slow and skill-bound. Shuang He and co-authors build BAMT-Net in a 2026 paper, a network that outlines lesions and labels them benign or malignant in one pass. The design leans on the fact that the two tasks inform each other.
A boundary-aware side task sharpens the outline work, while a guidance module feeds that shape sense into the labeling. An attention scheme blends space and channel cues to pick lesions out of busy tissue. The authors report clear gains on both tasks over split pipelines. They offer the system as a practical aid for faster, steadier breast screening.
Catalog record
| CREATOR | Shuang He, Dan Ji, Mingwei Ma, Xiang Pan, Juxiang Xu, Feng Liu |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 119 KB |
| VERSION | 1.0 |
| LICENSE | CC BY 4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |