ICHOR: A Robust Representation Learning Approach for ASL CBF Maps with Self-Supervised Masked Autoencoders

Xavier Beltran-Urbano, Yiran Li, Xinglin Zeng, Katie R. Jobson, Manuel Taso, Christopher A. Brown, David A. Wolk, Corey T. McMillan, Ilya M. Nashrallah, Paul A. Yushkevich, Ze Wang, John A. Detre, Sudipto Dolui · 2026

Read ICHOR: A Robust Representation Learning Approach for ASL CBF Maps with Self-Supervised Masked Autoencoders by Xavier Beltran-Urbano, Yiran Li, Xinglin Zeng, Katie R. Jobson, Manuel Taso, Christopher A. Brown, David A. Wolk, Corey T. McMillan, Ilya M. Nashrallah, Paul A. Yushkevich, Ze Wang, John A. Detre, Sudipto Dolui (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-2603-05247.txt.

Description

Xavier Beltran-Urbano and team present ICHOR, a self-supervised early training approach for arterial spin labeling blood flow maps using masked autoencoders. Perfusion MRI quantifies cerebral blood flow without exogenous contrast, but progress is limited by variable image quality, site, vendor, and rule gaps, and scarce labels. ICHOR pretrains a Vision Transformer by masked image building on 11,405 scans from 14 studies, among the largest arterial spin labeling datasets assembled. On three diagnostic sorting tasks and one map-quality regression task, it outperforms current neuroimaging self-supervised methods adapted to ASL.

Catalog record

CREATORXavier Beltran-Urbano, Yiran Li, Xinglin Zeng, Katie R. Jobson, Manuel Taso, Christopher A. Brown, David A. Wolk, Corey T. McMillan, Ilya M. Nashrallah, Paul A. Yushkevich, Ze Wang, John A. Detre, Sudipto Dolui
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE25048
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2603.05247

Browse the full library · About the Tower of Babel