Neural signatures of stream segregation: From childhood to adulthood

When faced with noisy environments, listeners perform auditory scene analysis, which allows them to parse the auditory target from concurrent interferences. Stream segregation involves organizing similar sound waves into a coherent stream, while distinguishing dissimilar acoustic components and attributing them to distinct sources. Two event-related potential components have been identified as "neural signatures" of stream segregation: the Object-Related Negativity (ORN) and the P400. Our study aims to examine (i) the maturation of neural and behavioural correlates of stream segregation and (ii) the development of the relationship between stream segregation and speech perception in noise. ORN/P400 were recorded while 8-23 year-olds performed an active stream segregation task based on temporal coherence. Participants also performed speech identification in noise tasks (behaviourally). Behavioral results indicate an improvement in both stream segregation and speech perception in noise from childhood to adulthood. Amplitude of the ORN (but not P400) decreases, and latency of both ORN and P400 decreases throughout development. Critically, P400 amplitude significantly predicts stream segregation performance. Overall, our results suggest that the neural mechanisms underlying stream segregation follow a prolonged maturation trajectory, and support the progressive maturation of auditory scene analysis and speech perception in noise. RESEARCH HIGHLIGHTS- Neurophysiological indices of stream segregation mature from childhood to adulthood - Neural indices of top-down attentive processing of complex auditory sequences predict stream segregation, irrespective of the listeners age - In adults (but not in children or adolescents), stream segregation predicts speech perception in noise - Stream segregation may thus operate as a bottleneck to speech in noise difficulties in children and adolescents