Latent Representation Learning of Cortical Atrophy using BERT‐based Foundation Model
Heeyun An, Jiyun Kim, Joon‐Kyung Seong, Wha Jin Lee · 2025
Read Latent Representation Learning of Cortical Atrophy using BERT‐based Foundation Model by Heeyun An, Jiyun Kim, Joon‐Kyung Seong, Wha Jin Lee (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-11716058.txt.
Description
Heeyun An and co-authors pretrain a brain-aging model on unlabeled scans in a 2025 neuroimaging paper. Cortical thickness maps reveal how neurodegeneration thins the brain, but every prediction task used to demand its own large labeled dataset and bespoke model.
Borrowing the BERT recipe from language modeling, the team fed thickness values from sixty-eight brain regions, measured in over ten thousand MRI scans, into a model that learns by guessing masked-out values. Once pretrained on thousands of unlabeled profiles, the foundation model needs only a thin output layer to tackle classification or staging. Context-aware thickness fingerprints are the product.
Catalog record
| CREATOR | Heeyun An, Jiyun Kim, Joon‐Kyung Seong, Wha Jin Lee |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 3 KB |
| VERSION | 1.0 |
| LICENSE | CC BY 4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |