Behavioral response to strong aversive stimuli: A neurodynamical model

In this paper a theoretical model of functioning of a neural circuit during a behavioral response has been proposed. A neural circuit can be thought of as a directed multigraph whose each vertex is a neuron and each edge is a synapse. It has been assumed in this paper that the behavior of such circuits is manifested through the collective behavior of neurons belonging to that circuit. Behavioral information of each neuron is contained in the coefficients of the fast Fourier transform (FFT) over the output spike train. Those coefficients form a vector in a multidimensional vector space. Behavioral dynamics of a neuronal network in response to strong aversive stimuli has been studied in a vector space in which a suitable pseudometric has been defined. The neurodynamical model of network behavior has been formulated in terms of existing memory, synaptic plasticity and feelings. The model has an analogy in classical electrostatics, by which the notion of force and potential energy has been introduced. Since the model takes input from each neuron in a network and produces a behavior as the output, it would be extremely difficult or may even be impossible to implement. But with the help of the model a possible explanation for an hitherto unexplained neurological observation in human brain has been offered. The model is compatible with a recent model of sequential behavioral dynamics. The model is based on electrophysiology, but its relevance to hemodynamics has been outlined.

💡 Research Summary

The paper proposes a novel neuro‑dynamical framework for describing how a neuronal network generates a behavioral response to a strong aversive stimulus. The authors first abstract a neural circuit as a directed multigraph, where vertices represent individual neurons and edges represent synaptic connections. Each neuron’s output spike train is transformed by a fast Fourier transform (FFT); the resulting spectral coefficients constitute a high‑dimensional “behavioral vector” that encodes the temporal firing pattern in the frequency domain.

A pseudometric is defined on the space of these vectors, not as a conventional Euclidean distance but as an analogue of electric potential difference. In this analogy, the existing memory of the network is modeled as a static potential field, while the aversive stimulus creates a potential gradient that exerts a “force” on the system. The force drives the network state toward a new configuration that minimizes a defined energy functional, mirroring the principle of electrostatic equilibrium.

Synaptic plasticity is interpreted as a redistribution of electric charge: changes in synaptic weights alter the shape of the potential field itself. Thus, the network’s memory (potential) is dynamically reshaped by plasticity. The authors further introduce “feelings” as an external electric field that biases the potential landscape. The same physical stimulus can therefore produce different behavioral vectors depending on the emotional state, providing a quantitative account of affect‑modulated decision making.

Although the model is mathematically intricate and would be extremely difficult to implement in a full‑scale simulation, the authors argue that it is compatible with existing sequential behavioral models such as Markov decision processes or reinforcement‑learning frameworks. By discretizing the continuous evolution of behavioral vectors into state transitions, the proposed formalism can be mapped onto conventional state‑space representations.

A further contribution is the linkage of the electrostatic analogy to hemodynamics. The authors suggest that changes in the electrical potential field correspond to metabolic demand, which in turn triggers localized increases in cerebral blood flow. This provides a conceptual bridge between electrophysiological signals and functional‑imaging observations, supporting a multimodal view of brain activity during aversive processing.

Finally, the model is applied to explain a previously puzzling neurological observation: prolonged post‑stimulus potential elevation in prefrontal regions after exposure to intense aversive cues. The authors propose that the interaction of pre‑existing memory (static potential), synaptic weight adjustments (charge redistribution), and emotional bias (external field) can generate the observed sustained potential. They outline experimental predictions—such as measurable shifts in spectral coefficients and correlated blood‑oxygen‑level‑dependent (BOLD) signals—that could validate the framework.

In summary, the paper offers a mathematically rigorous, physics‑inspired representation of neural circuit dynamics. By encoding spike trains as FFT‑derived vectors, defining a potential‑based pseudometric, and incorporating memory, plasticity, and affect as components of an electrostatic system, it provides a unified description of how strong aversive stimuli are transformed into coordinated behavioral outputs. The work bridges electrophysiology, computational modeling, and neuro‑vascular coupling, and suggests concrete avenues for empirical testing.