Determining the directions of information flow between populations of spiking neurons

Understanding the causal relationships between different cortical areas of the brain can help reveal the mechanisms of information processing and the interactions between these areas. Understanding how the brain works will help further the development of cognitive artificial intelligence systems. This paper investigates the possibility of determining the direction of information flow in the modeled populations of spiking neurons using pairwise Granger causality (GC) and partial directional coherence. V1 and V4 mutually predict each other, indicating that gamma rhythms are associated with feedforward communication and alpha/low-beta rhythms with feedback communication. The findings suggest that Granger causality and partial directional coherence are effective tools for determining causality between brain regions.