Integrating EEG phase-connectivity and behavior to study action perception in infancy

• In Event: Poster Session 05
  In Poster Session: PS 05 Topic - Biological Processes: Neuroscience and Genetics

Abstract

Despite the developmental importance of infants’ sensitivity to others’ goal-structured actions, relatively little is known about the neural underpinnings of this foundational social-cognitive skill. To date, nearly all studies of the neural correlates of action perception in infants have relied on a single measure of motor system dynamics: event-related desynchronization in the mu frequency band over central recording sites (Cuevas et al., 2014; Marshall & Meltzoff, 2011). This neural signature has proven useful in identifying motor system engagement, but it provides a limited window into the neural networks that support action processing. Here, we propose an innovative methodological approach to study the development of action perception by examining how infants’ motor competence shapes the synchronization of neural mechanisms involved in action observation. Nine- and twelve-month-old infants saw actions that were either familiar (grasping) or unfamiliar (pulling a cane) while their EEG signal was measured. At the end of the session, we coded infants’ latency to grasp toys or to pull a cane. Current source density (CSD) was applied to EEG data, and time-frequency and connectivity analyses were performed in CSD transformed data. Intersite phase-clustering (ISPC) across time between 104 electrodes was calculated for the infant alpha (6-9Hz) band and reorganized in 15 virtual electrode clusters. For each age group (9m; 12m) and condition (Grasp; Cane) we calculated participants’ whole brain connectivity as the median ISPC + 1SD of all electrode clusters at each time point (wb-ISPC). Connectivity between occipital and central areas (O-C) was also calculated as a measure of visuomotor coupling, with fronto-central (F-C) and parieto-central (P-C) as control cluster pairs (Debnath et al., 2019). We found greater wb-ISPC for 12-month-old infants than 9-month-old infants (F(1,34)= 4.67, p= 0.038), which suggests a link of wb-ISPC with maturation. Additionally, both age groups showed greater wb-ISPC during the anticipation of cane actions relative to grasp actions (F(1,34)= 5.68, p= 0.023). During this anticipatory period, we found an interaction between age and cluster pairs (F(1,68)= 3.42, p= 0.038): 12-month-olds, but not 9-month-olds, showed greater O-C ISPC relative to F-C and P-C ISPC. In a further analysis, we found that EEG connectivity during action anticipation predicted infants’ grasping competence: greater wb-ISPC predicted slower grasping (p= 0.013, R2= 0.12), whereas greater O-C ISPC predicted faster grasping (p= 0.0014, R2= 0.17). No effects or interactions were found for the other models (F-C; P-C) and no effects were found when relating connectivity to cane latency. These findings suggest that wb-ISPC is modulated by the familiarity of observed actions: infants showed more wb-ISPC when anticipating actions involving an unfamiliar tool. Additionally, for older infants O-C coupling was greater relative to other cluster pairs, suggesting a development of specific neural connections associated with action prediction and processing. Finally, our results suggest a relation between motor experience and brain connectivity. More competence in infants’ grasping was associated to enhanced visuomotor coupling during action anticipation, and less synchronization across other neural regions.

Authors

• Marc Colomer, University of Chicago
  Presenting Author

• Haerin Chung, University of Chicago
  Non-Presenting Author

• Marlene Meyer, Radboud Universiteit Nijmegen
  Non-Presenting Author

• Amanda Woodward, University of Chicago
  Non-Presenting Author