Isolating Nonlinear Independent Sources in fMRI with β \beta -TCVAE Models
2026
Read Isolating Nonlinear Independent Sources in fMRI with β \beta -TCVAE Models (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/arxiv-2605-16708.txt.
Description
Li and colleagues adapt a beta-TCVAE model, a kind of variational autoencoder that learns separate latent factors, to split nonlinear brain signals in fMRI data. Standard independent component analysis assumes linear mixing, which fails against the brain's nonlinear dynamics. Their model recovers real nonlinear spatial parts, including the default mode network and other core connectivity networks. Checks of functional network links show the learned structure captures coherent, clear brain patterns.
Catalog record
| YEAR | 2026 |
|---|---|
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 17753 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2605.16708