Limits...
Synaptic plasticity enables adaptive self-tuning critical networks.

Stepp N, Plenz D, Srinivasa N - PLoS Comput. Biol. (2015)

Bottom Line: We show that a combination of short- and long-term synaptic plasticity enables these networks to exhibit criticality in the face of intrinsic, i.e. self-sustained, asynchronous spiking.Brief external perturbations lead to adaptive, long-term modification of intrinsic network connectivity through long-term excitatory plasticity, whereas long-term inhibitory plasticity enables rapid self-tuning of the network back to a critical state.The critical state is characterized by a branching parameter oscillating around unity, a critical exponent close to -3/2 and a long tail distribution of a self-similarity parameter between 0.5 and 1.

View Article: PubMed Central - PubMed

Affiliation: Center for Neural and Emergent Systems, Information and System Sciences Lab, HRL Laboratories LLC, Malibu, California, United States of America.

ABSTRACT
During rest, the mammalian cortex displays spontaneous neural activity. Spiking of single neurons during rest has been described as irregular and asynchronous. In contrast, recent in vivo and in vitro population measures of spontaneous activity, using the LFP, EEG, MEG or fMRI suggest that the default state of the cortex is critical, manifested by spontaneous, scale-invariant, cascades of activity known as neuronal avalanches. Criticality keeps a network poised for optimal information processing, but this view seems to be difficult to reconcile with apparently irregular single neuron spiking. Here, we simulate a 10,000 neuron, deterministic, plastic network of spiking neurons. We show that a combination of short- and long-term synaptic plasticity enables these networks to exhibit criticality in the face of intrinsic, i.e. self-sustained, asynchronous spiking. Brief external perturbations lead to adaptive, long-term modification of intrinsic network connectivity through long-term excitatory plasticity, whereas long-term inhibitory plasticity enables rapid self-tuning of the network back to a critical state. The critical state is characterized by a branching parameter oscillating around unity, a critical exponent close to -3/2 and a long tail distribution of a self-similarity parameter between 0.5 and 1.

Show MeSH

Related in: MedlinePlus

Dependence of criticality and balance on Inhibitory STDP.A) The relationship between β and firing rate found for many simulation instances during a parameter search. The red square indicates the network described in Table 1. If inhibitory STDP is included, there is a region β ≈ [1.1,1.7] where relatively low firing rates are possible. Without inhibitory STDP, there is a critical value at β ≈ 1.05 that separates extremely high firing rates from networks that do not persist. B) A demonstration that if inhibitory STDP is turned off at 300 s, the avalanche power-law exponents fail to tune towards −1.5. Compare with Fig. 2a. C) The distribution of branching ratio over time and DFA α over neurons after the switch at 300 s. These show a decoupling of balance and criticality. D) Distributions of current balance after the switch at 300 s, showing that balance is retained even after I-STDP is switched off.
© Copyright Policy
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4295840&req=5

pcbi.1004043.g009: Dependence of criticality and balance on Inhibitory STDP.A) The relationship between β and firing rate found for many simulation instances during a parameter search. The red square indicates the network described in Table 1. If inhibitory STDP is included, there is a region β ≈ [1.1,1.7] where relatively low firing rates are possible. Without inhibitory STDP, there is a critical value at β ≈ 1.05 that separates extremely high firing rates from networks that do not persist. B) A demonstration that if inhibitory STDP is turned off at 300 s, the avalanche power-law exponents fail to tune towards −1.5. Compare with Fig. 2a. C) The distribution of branching ratio over time and DFA α over neurons after the switch at 300 s. These show a decoupling of balance and criticality. D) Distributions of current balance after the switch at 300 s, showing that balance is retained even after I-STDP is switched off.

Mentions: Plasticity of inhibitory synaptic connections is supported by recent experimental and theoretical findings [22, 34]. To judge the importance of inhibitory plasticity, the network was simulated both with and without inhibitory STDP. Fig. 9a shows firing rates achieved as a function of one of the varying network parameters, β (the E-STDP depression to potentiation ratio). In the presence of inhibitory STDP, there is a reasonably wide range of β ≈ [1.1, 1.7] for which the network can attain reasonable firing rates. In the absence of inhibitory STDP, small changes of β either saturate firing rate at 320 Hz or neurons are quiescent, and no intermediate regime can be established.


Synaptic plasticity enables adaptive self-tuning critical networks.

Stepp N, Plenz D, Srinivasa N - PLoS Comput. Biol. (2015)

Dependence of criticality and balance on Inhibitory STDP.A) The relationship between β and firing rate found for many simulation instances during a parameter search. The red square indicates the network described in Table 1. If inhibitory STDP is included, there is a region β ≈ [1.1,1.7] where relatively low firing rates are possible. Without inhibitory STDP, there is a critical value at β ≈ 1.05 that separates extremely high firing rates from networks that do not persist. B) A demonstration that if inhibitory STDP is turned off at 300 s, the avalanche power-law exponents fail to tune towards −1.5. Compare with Fig. 2a. C) The distribution of branching ratio over time and DFA α over neurons after the switch at 300 s. These show a decoupling of balance and criticality. D) Distributions of current balance after the switch at 300 s, showing that balance is retained even after I-STDP is switched off.
© Copyright Policy
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4295840&req=5

pcbi.1004043.g009: Dependence of criticality and balance on Inhibitory STDP.A) The relationship between β and firing rate found for many simulation instances during a parameter search. The red square indicates the network described in Table 1. If inhibitory STDP is included, there is a region β ≈ [1.1,1.7] where relatively low firing rates are possible. Without inhibitory STDP, there is a critical value at β ≈ 1.05 that separates extremely high firing rates from networks that do not persist. B) A demonstration that if inhibitory STDP is turned off at 300 s, the avalanche power-law exponents fail to tune towards −1.5. Compare with Fig. 2a. C) The distribution of branching ratio over time and DFA α over neurons after the switch at 300 s. These show a decoupling of balance and criticality. D) Distributions of current balance after the switch at 300 s, showing that balance is retained even after I-STDP is switched off.
Mentions: Plasticity of inhibitory synaptic connections is supported by recent experimental and theoretical findings [22, 34]. To judge the importance of inhibitory plasticity, the network was simulated both with and without inhibitory STDP. Fig. 9a shows firing rates achieved as a function of one of the varying network parameters, β (the E-STDP depression to potentiation ratio). In the presence of inhibitory STDP, there is a reasonably wide range of β ≈ [1.1, 1.7] for which the network can attain reasonable firing rates. In the absence of inhibitory STDP, small changes of β either saturate firing rate at 320 Hz or neurons are quiescent, and no intermediate regime can be established.

Bottom Line: We show that a combination of short- and long-term synaptic plasticity enables these networks to exhibit criticality in the face of intrinsic, i.e. self-sustained, asynchronous spiking.Brief external perturbations lead to adaptive, long-term modification of intrinsic network connectivity through long-term excitatory plasticity, whereas long-term inhibitory plasticity enables rapid self-tuning of the network back to a critical state.The critical state is characterized by a branching parameter oscillating around unity, a critical exponent close to -3/2 and a long tail distribution of a self-similarity parameter between 0.5 and 1.

View Article: PubMed Central - PubMed

Affiliation: Center for Neural and Emergent Systems, Information and System Sciences Lab, HRL Laboratories LLC, Malibu, California, United States of America.

ABSTRACT
During rest, the mammalian cortex displays spontaneous neural activity. Spiking of single neurons during rest has been described as irregular and asynchronous. In contrast, recent in vivo and in vitro population measures of spontaneous activity, using the LFP, EEG, MEG or fMRI suggest that the default state of the cortex is critical, manifested by spontaneous, scale-invariant, cascades of activity known as neuronal avalanches. Criticality keeps a network poised for optimal information processing, but this view seems to be difficult to reconcile with apparently irregular single neuron spiking. Here, we simulate a 10,000 neuron, deterministic, plastic network of spiking neurons. We show that a combination of short- and long-term synaptic plasticity enables these networks to exhibit criticality in the face of intrinsic, i.e. self-sustained, asynchronous spiking. Brief external perturbations lead to adaptive, long-term modification of intrinsic network connectivity through long-term excitatory plasticity, whereas long-term inhibitory plasticity enables rapid self-tuning of the network back to a critical state. The critical state is characterized by a branching parameter oscillating around unity, a critical exponent close to -3/2 and a long tail distribution of a self-similarity parameter between 0.5 and 1.

Show MeSH
Related in: MedlinePlus