Neuroadaptive Bayesian Optimisation to Predict an Individual Infant's Brain Response Towards Faces Resembling Parent or Stranger, 2022

This proof-of-principle study extends the novel experimental approach of Neuroadaptive Bayesian Optimisation (NBO) to infant EEG data to study individual infants' engagement with social stimuli. In particular, the Negative central event-related potential component was optimised across a range of familiar and nonfamiliar faces. Previous group-level research suggested an initial attentional preference for parent’s vs. stranger’s face around 6 months with a subsequent change towards enhanced attention to the stranger’s face. The present individualised NBO study included n=62 infants aged 5-12 months who were presented with faces linearly varying in similarity to parent’s face. Results showed lower-than-usual attrition rate, and an equal proportion of infants preferably attending to parent and stranger, with the individual’s probability of preferably attending to parent’s face increasing between 5 and 12 months but being unrelated to parent-reported social behaviour. This study proves the feasibility of the NBO approach with infant neurophysiological data to identify among a range of cues the one that maximally triggers social brain activity in the individual infant. Further, this study suggests that on the individual level, infants differ in whether they preferably attend to parent or stranger in the second half of the first year of life.