STD-Net: a spatio-temporal decoupling network for multiphasic liver lesion segmentation and characterization
Shaoliang Zhu, Mengjie Zou, Qijun Wu, Zheng Gong, Zhangnan Huang, Yan Zou, Tingting Tan, Yanwu You, Xiaofeng Dong, Honglin Luo · 2025
Read STD-Net: a spatio-temporal decoupling network for multiphasic liver lesion segmentation and characterization by Shaoliang Zhu, Mengjie Zou, Qijun Wu, Zheng Gong, Zhangnan Huang, Yan Zou, Tingting Tan, Yanwu You, Xiaofeng Dong, Honglin Luo (2025) 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-12775502.txt.
Description
Shaoliang Zhu and nine colleagues present STD-Net, a neural network for finding and characterizing liver tumors in multiphasic scans. Multiphasic CT and MRI capture how lesions enhance over time, a signature radiologists use for diagnosis. Most networks flatten the phases into channels and lose that temporal story.
STD-Net instead decouples the problem: a shared three-dimensional encoder extracts anatomy, and a transformer module models the sequence of contrast patterns, like arterial brightening followed by washout. On three public datasets it beat leading baselines in segmentation overlap and classification accuracy. Gains were most stable for small or low-contrast lesions. The authors propose spatio-temporal decoupling as a general pattern for dynamic medical imaging.
Catalog record
| CREATOR | Shaoliang Zhu, Mengjie Zou, Qijun Wu, Zheng Gong, Zhangnan Huang, Yan Zou, Tingting Tan, Yanwu You, Xiaofeng Dong, Honglin Luo |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 65 KB |
| VERSION | 1.0 |
| LICENSE | CC BY 4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |