| Abstract No.: | C-B3063 |
| Country: | Canada |
| | |
| Title: | MECHANISM OF ACTIVITY TRANSITIONS IN SPATIAL NEURAL NETWORKS |
| | |
| Authors/Affiliations: | 3 Elan Liss Ohayon; 3 Elan Liss Ohayon*; 3 W. McIntyre Burnham; 3 W. McIntyre Burnham; 2 Hon C. Kwan; 2 Hon C. Kwan; 1 Terrence J. Sejnowski; 1 Terrence J. Sejnowski; 1 Maxim Bazhenov; 1 Maxim Bazhenov;
1 Computational Neurobiology Laboratory, The Salk Institute, La Jolla, CA, USA; 2 Department of Physiology, University of Toronto, ON, Canada; 3 University of Toronto Epilepsy Research Program, Toronto, ON, Canada
|
| | |
| Content: | Onjectives: In this study we modeled persistent neural activity in laminar networks in order to identify fundamental factors that can trigger transitions in population dynamics.
Materials and methods: Simulations included spatial networks of spiking neurons with up to 10,000 neurons with separate inhibitory and excitatory populations. Networks had varying degrees of local and columnar connectivity. We examined how changes in neural density alter dynamics.
Results: Network models with heterogeneous connectivity showed various patterns of persistent activity including propagating waves. When activity of individual units was averaged over the population to simulate field recordings the network dynamics showed ongoing changes in both amplitude and spectral properties. These global patterns of network activity and the persistence of propagating waves were a function of both the degree of connectivity between neurons and initial activity conditions. Surprisingly, for certain connectivity densities, ongoing changes in the characteristics of network oscillations were present even in the absence of any further alteration to network structure or the intrinsic properties of the units.
Conclusions: It is often assumed that transitions in dynamics are due to (a) shifts in neural properties (b) changes to network structure (c) external input or (d) noise. The observation that intermittent transitions can take place in the absence of such factors suggests that we must look beyond these assumptions and on to the spatio-temporal features of neural dynamics. Specifically, transitions in signal features, especially when viewed at the population level, may be an indication of ongoing autonomous changes in spatial wave propagation. This finding could have important implications for the interpretation of field recordings. Lastly, the observations that changes in network architecture have effects on propensity and patterns of propagations could help explain changes in neural activity seen in conditions such as post-traumatic epilepsy, neurodegeneration and aging.
|
| | |
| Back |
|
