Discovering how emerging oscillatory dynamics interact with attention to shape memory representations in infancy - PROJECT ABSTRACT Cognitive development in infancy includes both learning to attend and attending to learn (Forest & Amso, in press). In other words, the emerging relationship between learning and attention is bidirectional, but the precise neural mechanisms which guide their reciprocal development are unknown. Attention can be conceptualized as an ongoing competition for neural resources within and between bottom-up and top-down signals, the winner of which is ‘attended’. In adults, EEG data indicate low-frequency, top-down oscillations originating in anterior brain regions modulate higher-frequency, bottom-up, more posterior oscillations via “phase-amplitude coupling” (PAC). PAC is thus a promising candidate for studying top-down/bottom-up competition in the infant brain. Critically, these oscillatory shifts have consequences for memory. This proposal is the first we know of to test the idea that attention and memory development are intertwined, such that PAC changes with age (Aim 1) and learning from past experience (Aim 2a), with cascading effects on infants’ visual selective attention and memory representations (Aim 2b). For Aim 1, we will chart PAC longitudinally from 3-24 months, as characterizing this balance across the first years of life offers novel insight into how attention is instantiated in the developing brain. Our pilot results support the hypothesis that theta-gamma PAC increases after 12mos, reflecting increasing contributions of top-down signal on posterior brain function. For Aim 2a we will further link the co-emergence of learning and attention by asking if prior experience quickly influences the balance of top-down/bottom-up neural signals. Infants (6-24mos) will complete a memory-guided-attention task during EEG. All trials show a target image amongst distractors. Some trials repeat, while others are new. We will compute PAC for each trial, and predict that PAC will dynamically increase over repeated trials, and that this increase will correlate with behavioral indices of memory-guided attention. To truly probe the co-evolving of PAC and infant learning, Aim 2b is to explain how PAC relates to developmental change in memory. In adults, stronger PAC leads to more consistent high-frequency activity, and high-frequency activity predicts memory specificity. Less consistent representations are associated with forming general memories in children and adults in other learning contexts. Thus, as PAC matures, bottom-up signals might become more consistent, and infants might remember increasingly specific details about their experiences. Following the memory guided-attention task, we will test infants’ specific and general memory for targets. Then, using spectral pattern similarity analyses, we will examine the consistency of the EEG signal across memory-guided attention trials. We predict infants who show less high-frequency neural similarity across repeated trials will show more general memory on the memory test. That is, as PAC increases, reflecting more top-down control, repeated-trial high-frequency activity will stabilize to support more specific memory. This project will mechanize the ways in which oscillatory dynamics, age, and experience co-evolve to support early learning, while providing me with unparalleled training in infant EEG and attention research.
