Aligning human subjects with short acquisition-time fMRI training data
Alexis Thual, Stanislas Dehaene, CEA, France; Bertrand Thirion, Inria, France
Posters 2 Poster
Pacific Ballroom H-O
Fri, 26 Aug, 19:30 - 21:30 Pacific Time (UTC -7)
Anatomical and functional inter-subject variability in the cortex is a major impediment to computing meaningful insights from cohorts of individuals. Recent techniques leveraging optimal transport have been shown to build alignments which greatly improve inter-subject correlation between fMRI contrast maps. However, datasets used to train these alignments are very time- and money-consuming to acquire. We replicate these experiments with much more affordable training datasets and show that they can already help derive meaningful alignments and significantly increase correlation between subjects.