Fast inter-subject alignment method for large datasets shows fine-grained cortical reorganisations
Alexis Thual,Huy Tran,Bertrand Thirion,Stanislas Dehaene · COSYNE 2022
Between-subject brain variability in shape and function is a major challenge to the definition of accurate brain models [1, 2]. It also obscures the comparison between species. Recently, precision mapping has started to provide data to ground the definition of accurate brain models [3]. Yet technology has been missing to identify correspondences between brains. Leveraging Optimal Transport (OT) methods [5, 6], we derive an algorithm, denoted as unbalanced fused Gromov Wasserstein (UFGW), to compute whole-brain mappings between subjects with minimal anatomical priors, and provide a fast GPU-bas
Thu, Mar 17, 2022