Abstracts Track 2020

Nr: 4

Does Functional Connectivity Reflect Anatomical Connectivity? A Simulation Study


Julien Bloch, Patrick Zhang, Eric Shea-Brown and Azadeh Yazdan-Shahmorad

Abstract: As the human brain is a complicated network of 100 billion neurons and 100 trillion synapses, one of its most defining characteristics is the internal organization of connections between neurons. Mapping this pipeline is a monumental task, given the daunting number of anatomical nodes and connections. Experiments attempting to calculate connectivity typically therefore quantify “functional connectivity,” which is a measure of large-scale statistical correlations in activity between brain regions. Despite the prevalence of functional connectivity metrics in neuroscience, there is no clear understanding of the relationship between functional connectivity and anatomical connectivity. In this work we develop a platform to characterize the relationship between the two forms of connectivity. Our platform consists of a large-scale (>100,000 neuron) simulation of virtual neurons with connectivity structure mimicking that of the cerebral cortex. This simulation also generates readouts of voltages at surface and intracortical locations within the simulation, analogous to how local-field potential (LFP) recordings are sampled from the cortex via electrodes. The model corresponds to a section of primate cortex of size 1.5mm x 1.5mm x 2.6mm, which is on the order of recent experiments quantifying functional connectivity between electrodes on a neural implant, such as between Utah array electrodes (Bloch 2019). The framework we use to develop this model is VERTEX (Tomsett 2015), and the connectivity structure is largely informed by (Binzegger 2004). Our unique platform allows us to calculate functional connectivity metrics from the recorded LFP signals and compare them to the virtual anatomical connectivity specified by the model. We first compute several measures of anatomical connectivity strength between the distinct virtual electrode locations. We then compute standard functional connectivity metrics such as coherence and Granger causality from the LFP signals collected from the virtual electrodes, and use correlational analysis to investigate their relationship. Functional connectivity is a contentious topic in neuroscience. Although studies investigating functional connectivity are widespread, its synaptic underpinnings remain mysterious. In this work we develop a state-of-the-art computational platform to calculate functional and anatomical connectivity in a biorealistic virtual brain, and characterize their relationship to a degree that until now has been impossible. Binzegger, T., Douglas, R. J., & Martin, K. A. C. (2004). A quantitative map of the circuit of cat primary visual cortex. Journal of Neuroscience, 24(39), 8441–8453. Bloch, J. A., Khateeb, K., Silversmith, et. al. (2019). Cortical Stimulation Induces Network-Wide Coherence Change in Non-Human Primate Somatosensory Cortex. Proceedings of the Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBS, 6446–6449. Tomsett, R. J., Ainsworth, M., Thiele, et. al. (2015). Virtual Electrode Recording Tool for EXtracellular potentials (VERTEX): comparing multi-electrode recordings from simulated and biological mammalian cortical tissue. Brain Structure and Function, 220(4), 2333–2353.