Skull stripping with purely synthetic data Journal: arxiv
Jong Sung Park, Juhyung Ha, Siddhesh Thakur, Alexandra Badea, Spyridon Bakas, Eleftherios Garyfallidis · 2025
Read Skull stripping with purely synthetic data Journal: arxiv by Jong Sung Park, Juhyung Ha, Siddhesh Thakur, Alexandra Badea, Spyridon Bakas, Eleftherios Garyfallidis (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-2505-07159.txt.
Description
Jong Sung Park and five co-authors present a way to strip the skull from brain scans without ever training on real brain images. Skull stripping, which removes bone and other non-brain tissue, is a first step in much scan work. Their 2025 paper notes that many past tools were built for narrow cases, and a broadly reliable method is still missing.
Their method, called PUMBA, trains a model only on purely synthetic images and labels made for the task. No real scans and no built-in facts about brain anatomy guide the training. Even so, the model matched the accuracy of rival tools on scans of several kinds, across species, and in cases with disease present. The authors offer the work as a fresh path for other scan tasks that need models able to travel well.
Catalog record
| CREATOR | Jong Sung Park, Juhyung Ha, Siddhesh Thakur, Alexandra Badea, Spyridon Bakas, Eleftherios Garyfallidis |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 25 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2505-07159