Bridging Structure, Dynamics, and Computation in Brain Networks

The brain network has an exquisite ability to process sensory information robustly and efficiently. The complex connectivity structure and rich neuronal dynamics in the brain network are believed to underlie such optimal coding capability. In this talk, I will introduce two different approaches investigating how the brain network structure, dynamics, and computational strategies shape each other.

    The first part of the talk will focus on a neural circuitry composed of an intermediate visual cortical area V4 and its efferent prefrontal cortex. Specifically, I will discuss how its structure and dynamics promote robust sensory computation using predictive coding— a theory of cortical computation which has been proposed as a method to create efficient neural codes. The primate visual system recognizes partially occluded objects in natural scenes without difficulty, but the neural basis underlying such computational capability is poorly understood. Recent experimental results from primate area V4, an intermediate stage in the ventral visual pathway, suggests that the feedback from higher cortical areas may be important for maintaining robust shape selective activities under partial occlusion. We implement a predictive coding model of V4 and prefrontal cortex to investigate possible computational roles of feedback signals in the visual system and their potential significance in object recognition.

 In the second part of the talk, I will introduce a more general problem of linking structure and dynamics, by discussing how structural characteristics of the mesoscopic mouse whole-brain connectome contribute to unique network dynamics. Using a network of phase oscillators whose connectivity is defined by viral tracing experiments, we show that complexities added to the spatially embedded whole-brain connectome by idiosyncratic long-range connections, enable rapid transitions between local and global synchronizations.  These results, therefore, demonstrate how structure and dynamics of neuronal networks may shape computation and function in the brain, and vice versa.