Boosting of neural circuit chaos at the onset of collective oscillations

Agostina Palmigiano, Rainer Engelken, Fred Wolf

Abstract

Neuronal spiking activity in cortical circuits is often temporally structured by collective rhythms. Rhythmic activity has been hypothesized to regulate temporal coding and to mediate the flexible routing of information flow across the cortex. Spiking neuronal circuits, however, are non-linear systems that, through chaotic dynamics, can amplify insignificant microscopic fluctuations into network-scale response variability. In nonlinear systems in general, rhythmic oscillatory drive can induce chaotic behavior or boost the intensity of chaos. Thus, neuronal oscillations could rather disrupt than facilitate cortical coding functions by flooding the finite population bandwidth with chaotically-boosted noise. Here we tackle a fundamental mathematical challenge to characterize the dynamics on the attractor of effectively delayed network models. We find that delays introduce a transition to collective oscillations, below which ergodic theory measures have a stereotypical dependence on the delay so far only described in scalar systems and low-dimensional maps. We demonstrate that the emergence of internally generated oscillations induces a complete dynamical reconfiguration, by increasing the dimensionality of the chaotic attractor, the speed at which nearby trajectories separate from one another, and the rate at which the network produces entropy. We find that periodic input drive leads to a dramatic increase of chaotic measures at a the resonance frequency of the recurrent network. However, transient oscillatory input only has a moderate role on the collective dynamics. Our results suggest that simple temporal dynamics of the mean activity can have a profound effect on the structure of the spiking patterns and therefore on the information processing capability of neuronal networks.

Introduction: The Paradox of Oscillations

In neuroscience, collective rhythms such as alpha or gamma waves are traditionally viewed as organizational backbones for neural communication. These rhythms are thought to facilitate the routing of information across the cortex by synchronizing firing patterns. However, neural circuits are fundamentally nonlinear systems. In many physical contexts, adding a periodic drive to a nonlinear system does not simplify its behavior; instead, it can trigger or "boost" internal chaos.

This raises a critical question: Do neuronal oscillations quench noise to facilitate coding, or do they disrupt it by flooding the network with chaotically-boosted noise? Answering this has been historically difficult due to the mathematical complexity of synaptic delays. Systems with delays are technically infinite-dimensional, as their current state depends on the entire history of the delay period, making the calculation of chaos measures like Lyapunov exponents (LEs) notoriously difficult.

Methods: Mapping Infinite Dimensions to Finite Space

To overcome these mathematical hurdles, the authors developed a novel "two-stage" neuronal model. Rather than treating synaptic delay as a simple time-lag, they modeled each neuron as a soma (cell body) linked to a Single Compartment Axon (SCA).

Asynchronous state without synaptic delays (delta=0), where the system dimensionality is proportional to the number of neurons N.

By treating the axon as a separate unit that relays spikes after a specific duration, the researchers mapped the infinite-dimensional delayed system onto a tractable finite-dimensional system with $2N$ dimensions. This innovation allowed for the semi-analytical computation of the full spectrum of Lyapunov exponents, which in turn enabled the derivation of the Kaplan-Yorke dimension (measuring attractor complexity) and metric entropy (measuring noise production rate).

Results: The Chaos Boost and Dimensionality Expansion

The study reveals that as synaptic delays increase, the network undergoes a "dynamical reconfiguration." At a critical point known as a Hopf bifurcation, steady asynchronous activity breaks into collective oscillations (typically in the beta range, ~20 Hz).

The onset of collective oscillations with synaptic delays (delta != 0) leads to a reconfiguration of the network's chaotic attractor.

At this transition, both the maximum Lyapunov exponent and the attractor dimension spike significantly. Before oscillations emerge, increasing the delay expands the number of "directions" in which the system is chaotic, even if the total entropy remains stable. Once oscillations begin, both the intensity and dimensionality of chaos are boosted.

Key findings include:

  • Extensive Chaos: Chaos in these networks is "extensive," with attractor dimension and entropy growing linearly with the number of neurons ($N$).
  • Resonance Effects: External periodic signals dramatically intensify chaos if their frequency matches the network's natural resonance.
  • The Role of Heterogeneity: In uniform networks, long delays can lead to pathological perfect synchrony. However, realistic cell-to-cell heterogeneity prevents this collapse, maintaining the network's chaotic flexibility.

Discussion: Transience as a Safeguard

While a constant periodic drive can boost chaos by up to 40%, the authors found that realistic transient oscillatory bursts—the type most common in the brain—only have a moderate effect (~5% increase). This suggests that the brain may utilize transient rhythms to gain the benefits of temporal organization without incurring the full "noise penalty" of intensified chaos. These results indicate that the temporal dynamics of mean activity have a profound effect on spiking patterns and, consequently, the information-processing capacity of the cortex.

Summary updated

Abstract

Neuronal spiking activity in cortical circuits is often temporally structured by collective rhythms. Rhythmic activity has been hypothesized to regulate temporal coding and to mediate the flexible routing of information flow across the cortex. Spiking neuronal circuits, however, are non-linear systems that, through chaotic dynamics, can amplify insignificant microscopic fluctuations into network-scale response variability. In nonlinear systems in general, rhythmic oscillatory drive can induce chaotic behavior or boost the intensity of chaos. Thus, neuronal oscillations could rather disrupt than facilitate cortical coding functions by flooding the finite population bandwidth with chaotically-boosted noise. Here we tackle a fundamental mathematical challenge to characterize the dynamics on the attractor of effectively delayed network models. We find that delays introduce a transition to collective oscillations, below which ergodic theory measures have a stereotypical dependence on the delay so far only described in scalar systems and low-dimensional maps. We demonstrate that the emergence of internally generated oscillations induces a complete dynamical reconfiguration, by increasing the dimensionality of the chaotic attractor, the speed at which nearby trajectories separate from one another, and the rate at which the network produces entropy. We find that periodic input drive leads to a dramatic increase of chaotic measures at a the resonance frequency of the recurrent network. However, transient oscillatory input only has a moderate role on the collective dynamics. Our results suggest that simple temporal dynamics of the mean activity can have a profound effect on the structure of the spiking patterns and therefore on the information processing capability of neuronal networks.
BibTeX
@misc{palmigiano2023boosting,
  author = {Agostina Palmigiano and Rainer Engelken and Fred Wolf},
  title = {{Boosting of neural circuit chaos at the onset of collective oscillations}},
  year = {2023},
  doi = {10.7554/elife.90378},
}
Figure