SST-DUNet: Automated preclinical functional MRI skull stripping using Smart Swin Transformer and Dense UNet
Sima Soltanpour, Rachel Utama, Arnold Chang, Md Taufiq Nasseef, Dan Madularu, Praveen Kulkarni, Craig Ferris, Chris Joslin · 2025
Read SST-DUNet: Automated preclinical functional MRI skull stripping using Smart Swin Transformer and Dense UNet by Sima Soltanpour, Rachel Utama, Arnold Chang, Md Taufiq Nasseef, Dan Madularu, Praveen Kulkarni, Craig Ferris, Chris Joslin (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/arxiv-2504-19937.txt.
Description
Sima Soltanpour and co-authors automate a tedious step in brain scan studies of rats in a 2025 paper. Before brain activity can be read, the skull must be stripped from each scan image. Done by hand, that step is slow and varies by person, and animal scans add low detail and odd slice sizes.
Their model blends two network designs, one that widens features and one that attends to brain structure in channels. A mixed loss function handles the skew between brain and non-brain pixels. Scores on three in-house data sets neared 99 percent match with hand work. Later brain analyses from the auto-stripped scans closely matched those from manual work.
Catalog record
| CREATOR | Sima Soltanpour, Rachel Utama, Arnold Chang, Md Taufiq Nasseef, Dan Madularu, Praveen Kulkarni, Craig Ferris, Chris Joslin |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 53 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2504-19937