Disentangling the functional roles of premotor-motor pathways in automatic imitation: A combined network-based transcranial stimulation and drift diffusion modeling approach.

Humans have an automatic tendency to imitate others' actions, a process facilitated by the action observation network (AON). While motor nodes of the AON, such as the ventral premotor cortex (PMv) and the supplementary motor area (SMA), are engaged during automatic imitation, the distinct roles of their projections to the primary motor cortex (M1) remain poorly understood. Here, we investigate the plasticity and functional role of PMv-to-M1 and SMA-to-M1 pathways in healthy humans of either sex. We used a combination of cortico-cortical paired associative stimulation (ccPAS) to modulate cortical connectivity strength, and drift diffusion modeling (DDM) to study the impact of ccPAS on the latent cognitive processes underlying automatic imitation. Our results show that manipulating PMv-to-M1 connectivity increases the baseline tendency to imitate actions, shifting the response toward or away from an imitative response when connectivity in this circuit is enhanced or hindered, respectively. Conversely, strengthening SMA-to-M1 connectivity does not affect this bias but improves contextual information integration, facilitating task-appropriate behavior, reflected by the drift rate parameter. These findings demonstrate a double dissociation in the functional roles of PMv-to-M1 and SMA-to-M1 pathways: the former pathway drives the automatic imitation bias, while the latter modulates the integration of contextual information to regulate imitation. By combining network-based brain stimulation with advanced behavioral analysis, this study provides causal evidence for the distinct cognitive functions supported by the PMv-to-M1 and SMA-to-M1 pathways in the facilitation and regulation of automatic imitation. Our findings offer insights into the neural mechanisms governing imitation and its context-dependent modulation.Significance statement Humans automatically imitate others' actions, a process supported by fronto-parietal brain regions. However, the specific roles of the ventral premotor to primary motor cortex (PMv-to-M1) and supplementary motor area to primary motor cortex (SMA-to-M1) pathways are unclear. Using cortico-cortical paired associative stimulation (ccPAS) and drift diffusion modeling (DDM), we demonstrate a double dissociation in the functions of these networks. PMv-to-M1 connectivity determines the baseline tendency to imitate (i.e., starting point parameter), either increasing or reducing automatic imitation. In contrast, SMA-to-M1 connectivity enhances integration of contextual information, promoting task-appropriate responses (i.e., drift rate parameter). These findings highlight distinct contributions of PMv-to-M1 and SMA-to-M1 pathways to imitation and its regulation, advancing our understanding of the neural mechanisms underlying social behavior.