Stereotyped spatiotemporal dynamics of spontaneous activity in visual cortex prior to eye-opening.

Understanding the temporal dynamics underlying the network structure in early development is critical for understanding network function and plasticity. By imaging hours of spontaneous cortical activity, we show strong evidence that the vast majority of spontaneous neural activity is dynamic with repeated and complex spatiotemporal patterns with stereotyped structure across hours. This suggests the potential for Hebbian learning in the development and refinement of functional visual representations. We also find that frequently occurring spatial activity patterns are predictive of subsequent activity for up to one second, which may indicate attractor dynamics in spontaneous activity. Our findings characterize key features of the temporal structure of spontaneous activity in visual cortex early in development and deepen our understanding of developing neural networks.