Functional Connectivity of the Brain: Reconstruction from Static and Dynamic Data

Abstract

The central nervous system is a single complex network connecting each neuron through a number of synaptic connections. However, only a small fraction of the total connections functionally link neurons together. If the smallest multineuronal architecture within which functional links are established constitutes circuitry, then what are the basic operating principles of these circuitries from which we can understand both the composition and the dynamics of the larger networks? We argue that a finite class of circuitries, the “basic circuitries,” can be identified as repeating structural motifs tightly associated with specific dynamics. Functional circuitries, however, cannot be derived from the static architecture simply because they do not obey structural borders. Fortunately, since the constituent neurons do act in synergy, we can infer from the dynamics the minimal structural conditions that constitute a circuitry. In this chapter, instead of giving a precise definition of the “basic circuitry,” we outline a set of methods that may elucidate such a definition. We argue that since the concept of circuitry incorporates both dynamic and static features, understanding can be achieved through combining the structural and dynamic aspects of the available data. We review methods of extracting functional information from static data first. Next, we review methods of extracting structural information from dynamic data. Ideally, these two approaches should converge and define circuitry based on the fragile concept of functional connectivity.