This title appears in the Scientific Report :
2018
A mesoscopic layered cortical network model for spiking activity and local field potentials
A mesoscopic layered cortical network model for spiking activity and local field potentials
Extracellular recordings with chronic or acute implants of multi-electrode arrays provide simultaneous access to population signals such as local field potentials (LFPs) and the spiking activity of hundreds or more individual neurons across several square millimeters of cortical tissue (Riehle et al...
Saved in:
Personal Name(s): | Senk, Johanna (Corresponding author) |
---|---|
Hagen, Espen / van Albada, Sacha / Diesmann, Markus | |
Contributing Institute: |
Jara-Institut Brain structure-function relationships; INM-10 Computational and Systems Neuroscience; IAS-6 Computational and Systems Neuroscience; INM-6 |
Imprint: |
2018
|
Conference: | Bernstein Conference 2018, Berlin (Germany), 2018-09-26 - 2018-09-28 |
Document Type: |
Abstract |
Research Program: |
Brain-Scale Simulations Supercomputing and Modelling for the Human Brain Human Brain Project Specific Grant Agreement 2 Human Brain Project Specific Grant Agreement 1 Connectivity and Activity Doktorand ohne besondere Förderung |
Publikationsportal JuSER |
Extracellular recordings with chronic or acute implants of multi-electrode arrays provide simultaneous access to population signals such as local field potentials (LFPs) and the spiking activity of hundreds or more individual neurons across several square millimeters of cortical tissue (Riehle et al., 2013). Signal predictions from multiscale computational models covering similar spatial dimensions help to interpret experimental data and expose the mechanisms underlying observed spatiotemporal activity patterns in cortex (Voges and Perrinet, 2012; Muller et al., 2018; Denker et al., 2018). Potjans and Diesmann (2014) modeled the local microcircuit under 1 mm2 of cortical surface area as a layered spiking point-neuron network integrating layer- and neuron-type specific connectivity data. From the simulated spiking activity LFPs can be computed using electrostatic forward predictions based on multicompartment models (Hagen, Dahmen et al., 2016). Here, we extend this microcircuit model to 4×4 mm2 as illustrated in Figure 1, to match the area covered by multi-electrode arrays in use today. The upscaling preserves the realistic neuron densities of the original model, and introduces distance-dependent connection probabilities and conduction delays. A sparsity of detailed experimental data on distance-dependent connectivity yields uncertainty in model parameters, which we address by testing parameter combinations within biologically plausible bounds. All network simulations are carried out with NEST (http://www.nest-simulator.org; Gewaltig and Diesmann, 2007) and LFPs are constructed using LFPy (http://lfpy.readthedocs.io) with NEURON (https://neuron.yale.edu) as back end. We find that the upscaled model preserves the overall spiking statistics of the original model with asynchronous irregular spiking across populations and weak pairwise spike-train correlations in line with experimental observations (Ecker et al., 2010). In contrast to the weak spike-train correlations, the spatial correlation of LFP signals is strong and distance-dependent, also matching experimental findings (Nauhaus et al., 2009). |