Abstract
Building object representations is crucial for understanding the visual world, but it is not yet understood how infants start to form these representations. In adults, theta power is higher during presentation of stimuli that were later remembered, compared to those later forgotten (Friese et al., 2013), and the coupling between theta phase and gamma amplitude has been shown to be responsible for binding perceptual features to form representations. Theta-gamma coupling has been observed, for example, during the formation of visual associations (Köster, Finger, Graetz, Kater & Gruber, 2018), and was again higher for remembered than forgotten stimuli. Theta-gamma coupling has not yet been examined in infants due to methodological difficulties, such as contamination of the EEG signal by microsaccades (Köster, 2016) and low signal-to-noise ratio (Hoehl & Wahl, 2012). Since infants are in the process of building novel representations of their environment all the time, they are well-suited to show how early representations can begin to form. The current study introduces tools to solve these limitations and examines the neuronal processes related to forming new representations in 48 infants (6;12 – 8;04 months old). At the beginning of the experiment, infants were familiarised with four stimuli. They then saw interleaved trials showing either a familiar object or an entirely novel object. EEG was recorded continuously and wavelet transformed to time-frequency space. We expected that infants would show higher theta power for novel compared to familiar stimuli, and that theta-gamma coupling would be higher for novel compared to familiar stimuli. Our results show increased parietal and frontal theta for novel compared to familiar objects, in line with theta’s role in building semantic conceptual networks. We were successful in removing microsaccade artifacts from the infant EEG and will further present the data on the cross-frequency coupling between theta phase and gamma amplitude.