Abstract
ABSTRACT
Motor responses are influenced by inferred predictability, but it remains unclear how uncertainty, or its inverse, precision, is encoded within motor control networks. In this study, we combined computational modeling of reaching movement response speed with functional magnetic resonance imaging (fMRI). We used a cueing paradigm with changing cue predictability (~50%, ~70%, and ~90%) to test how predictive inference about left/right reaching responses affected cortical activity and network connectivity in 28 healthy young adults. Behavioral results showed that the response speed differences between invalid and valid cueing increased with a higher model‐derived probability that the cue will be valid. The fMRI results revealed that activity in the bilateral intraparietal sulcus (IPS), the caudal part of dorsal premotor cortex (cPMd), and the cerebellum increased with higher uncertainty during valid trials. Moreover, the activation in the premotor cortex, bilateral inferior parietal lobe (IPL), IPS, and temporoparietal cortex was enhanced when reaching to invalidly cued targets (i.e., during motor reprogramming). Dynamic causal modeling revealed that uncertainty increased the connectivity from left to right IPS, from left IPS to cPMd, and from cPMd to primary motor cortex (M1) during valid trials. Motor reprogramming during invalid trials increased the coupling between bilateral IPL and strengthened pathways from IPL to M1 and from IPS to PMd, while decreasing connectivity from right to left IPS. These findings provide the first evidence of uncertainty‐dependent network dynamics during reaching movements in the human brain and shed light on distinct neural mechanisms underlying precision‐dependent motor control and motor reprogramming, with potential implications for psychiatric and neurological disorders characterized by aberrant precision control.