Data-Driven Mapping Between Functional Connectomes Using Optimal Transport

Functional connectomes derived from functional magnetic resonance imaging have long been used to understand the organization of brain. Nevertheless, a connectome is intrinsically linked atlas create it. In other words, generated one different in scale and resolution compared another atlas. Being able map results between atlases without additional pre-processing crucial step improving interpretation generalization studies that use atlases. Here, we optimal transport, powerful mathematical technique, find an optimum mapping two This then transform time series order reconstruct connectome. We validate our approach by comparing transformed against their “gold-standard” counterparts (i.e., directly atlas) demonstrate utility applying these predictive models based on show are significantly similar maintain individual differences brain-behavior associations, demonstrating both validity its downstream analyses. Overall, promising avenue increase connectome-based across