Globally conditioned causality in estimating directed brain-heart interactions through joint MRI and RR series analysis

We used 7T fMRI with simultaneous physiological signals acquisitions to investigate the causal interactions from resting state brain activity to autonomic nervous system (ANS) outflow as quantified through a probabilistic heartbeat model. Given the highly redundant nature of brain-derived signals, we compare the results of traditional bivariate Granger Causality (GC) to a globally conditioned approach which evaluates the additional influence of each brain region on ANS activity while factoring out effects concomitantly mediated by other brain regions. The bivariate approach results in an unrealistically large number of spurious causal brain-heart links. In contrast, using the globally conditioned approac, we demonstrate the existence of significant selective causal links between cortical/subcortical brain regions and ANS outflow for sympathetic and parasympathetic modulation as well as sympathovagal balance, with a prominent involvement of frontal, parietal, and cerebellar regions and Sensory Motor, Default Mode, Left and Right executive networks. Provided proper conditioning is employed to eliminate spurious causalities, 7T functional imaging coupled with physiological signal acquisition and GC analysis is able to quantify directed brain-heart interactions reflecting central modulation of ANS outflow.