Anatomic connections between brain areas affect information flow between neuronal circuits. However, such structural connectivity does not coincide with effective connectivity related to the more elusive question “Which areas cause the present activity of which others?”. Indeed effective connectivity depends flexibly on contexts and tasks and must be reconfigurable even when the underlying structural connectivity is fixed. Recently, computational mean-field models of whole-brain networks have reproduced with remarkable accuracy functional interactions in the so-called “resting state”, showing that they emerge spontaneously from the interplay between thalamocortical structure, interaction delays and noise-driven local dynamics.
Beyond the resting state, we will investigate how sensory- or cognitive-driven biases modulate the macro-scale dynamics of a simulated brain. Beyond mean-field approaches, we will model selected brain areas at a micro-scale level of detail, in order to describe correlations in their spiking activity and analyze how they are reorganized by changes in brain state.
More specifically, we will focus on functional interactions between brain areas belonging to the dorsal and ventral attention systems, known to be determinant for the initiation, the maintenance and the reorienting of selective attention. In this context, we will explore how the self-organization of neural activity controls the balance between top-down and bottom-up inter-areal influences, in different attentional conditions. We will then study how emergent dynamic patterns of multi-frequency phase-coherence enable flexible routing of information encoded in spiking activity.
This work will profit of the stimulating environment offered by the Systems Neuroscience Institute in Marseille, in charge of the development of “the Virtual Brain”, a high-performance-computing platform for realistic and potentially clinically-relevant “virtual imaging” experiments.
Call for proposal
See other projects for this call