Inter-areal balanced amplification enhances signal propagation in a large-scale circuit model of the primate cortex

Abstract

Reliable signal transmission represents a fundamental challenge for cortical systems, which display a wide range of weights of feedforward and feedback connections among heterogeneous areas. We re-examine the question of signal transmission across the cortex in network models based on recently available mesoscopic, directed-and weighted-inter-areal connectivity data of the macaque cortex. Our findings reveal that, in contrast to feed-forward propagation models, the presence of long-range excitatory feedback projections could compromise stable signal propagation. Using population rate models as well as a spiking network model, we find that effective signal propagation can be accomplished by balanced amplification across cortical areas while ensuring dynamical stability. Moreover, the activation of prefrontal cortex in our model requires the input strength to exceed a threshold, in support of the ignition model of conscious processing, demonstrating our model as an anatomically-realistic platform for investigations of the global primate cortex dynamics.In computational neuroscience, there is a dearth of knowledge about multi-regional brain circuits. New questions that are not crucial for understanding local circuits arise when we investigate how a large-scale brain system works. In particular, reliable signal propagation is a prerequisite for information processing in a hierarchically organized cortical system. A number of studies have been devoted to signal propagation from area to area in the mammalian cortex [1][2][3][4][5][6][7][8]. It was found that a major challenge is to ensure stable transmission, with the signal undergoing neither successive attenuation nor amplification as it travels across multiple areas in a hierarchy. In spite of insights these studies provided, virtually all previous works did not incorporate data-constrained cortical connectivity, and made unrealistic assumptions -for instance, areas are considered identical, network architecture is strictly feedforward, and connection weights are the same at all stages of the hierarchy. Here we argue that achieving stable signal propagation becomes even more challenging with the inclusion of more realistic network architecture and connectivity. Mechanisms that improve signal propagation in simpler models may no longer work for more biological models.Inter-areal cortical network is highly recurrent, abundant with feedback loops [9]. These rich feedback connections pose the risk of destabilizing the system through reverberation as the signal is transmitted across areas. Local cortical circuits are strongly recurrently connected [10], further contributing to system instability. Therefore, mechanisms that improve signal propagation in feedforward networks may quickly lead to instability in a brain-like network. Building such a model requires quantitative anatomical data. Recently, detailed mesoscopic connectivity data has become available for both macaque monkey [9][10][11] and mouse [12,13]. This data delineates a complex inter-areal cortical ...

Task-evoked activity quenches neural correlations and variability across cortical areas

“…We found that the population transfer function approximated a sigmoid activation function ( Figure 5B). Approximating the mean-field transfer function of a cortical area allowed us to focus our modeling efforts on simplified excitatory networks across large cortical areas, since most inter-area cortical networks rely on excitatory connectivity (Joglekar et al, 2018) . The population spike rate (excitatory cells only) subject to inhibitory regulation.…”

Mechanisms of distributed working memory in a large-scale network of macaque neocortex

“…1b) (Wang, 2001;Wong and Wang, 2006). In addition, there is a macroscopic gradient of synaptic excitation (Chaudhuri et al, 2015;Joglekar et al, 2018; Extended Data Fig. 2), namely the number of spines, loci of excitatory synapses, per pyramidal cell (Elston 2007) was used as a proxy for the strength of recurrent and long-range excitation that increases along the cortical hierarchy (Felleman and van Essen, 1991;Markov et al, 2014; Fig.…”

Systematic Integration of Structural and Functional Data into Multi-Scale Models of Mouse Primary Visual Cortex

“…The variants were constrained by a plethora of experimental data: the representation of the individual cells and their firing behavior in response to somatic current injections, LGN filters, thalamocortical connectivity, recurrent connectivity, and activity patterns observed in vivo. This work continues the trend of developing increasingly more sophisticated models of cortical circuits in general (e.g., Traub et al, 2005;Zhu, Shelley and Shapley, 2009;Potjans and Diesmann, 2014;Markram et al, 2015;Joglekar et al, 2018;Schmidt et al, 2018) and visual cortex in particular (Wehmeier et al, 1989;Troyer et al, 1998;Zemel and Sejnowski, 1998;Krukowski and Miller, 2001;Arkhipov et al, 2018;Antolík et al, 2019). Our main goal was to integrate existing and, especially, emerging multi-modal experimental datasets describing the structure and in vivo activity of cortical circuits into biologically realistic network models.…”

“…Simulating cortical circuits has a long history (e.g. (Wehmeier et al, 1989;Zemel and Sejnowski, 1998;Troyer et al, 1998;Krukowski and Miller, 2001;Traub et al, 2005;Zhu, Shelley and Shapley, 2009;Potjans and Diesmann, 2014;Markram et al, 2015;Arkhipov et al, 2018;Joglekar et al, 2018;Schmidt et al, 2018;Antolík et al, 2019;Schwalger and Chizhov, 2019)), with models incrementally building upon their predecessors. The simulations described here are a further instance of this evolution toward digital simulacra that predict new experiments, are insightful, and ever more faithful to the vast complexity of cortical tissue, in particular its heterogeneous neuronal cell classes, connections, and in vivo activity.…”