Summary:
Using a compact mathematical rate model of the primary visual cortex, computational neuroscientists demonstrate that two intrinsic feedback mechanisms—fast inhibitory recruitment and slower homeostatic regulation—can constrain chaotic activity and reduce neural variability by about 93%. These results offer a principled framework to resolve a core sensory paradox: how the visual system stays adaptable enough to detect a changing environment while avoiding runaway chaotic states that would corrupt perception.
Key Facts:
- Feedback cuts variability by ~93%: When biologically inspired cortical feedback loops (rapid interneuron inhibition and slow homeostatic drive) were added to a chaotic network model, variance in excitatory activity fell from 0.325 to 0.024 without eliminating the circuit’s capacity for flexible responses.
- Minimal three-population architecture: Instead of simulating millions of cells, the study’s E-I-M rate model reproduces key features of primate V1 using three interacting populations: excitatory pyramidal cells (E), fast inhibitory PV+ interneurons (I), and a slower modulatory drive (M).
- Precision maintained at the edge of chaos: Moderately irregular dynamics did not impair stimulus processing; virtual neurons matched macaque orientation selectivity index (OSI) ranges (approximately 0.31–0.38) and, under controlled near-threshold dynamics, showed slightly sharper orientation discrimination.
Source: International Centre for Translational Eye Research (ICTER) / Institute of Physical Chemistry, Polish Academy of Sciences
Every time we shift our gaze, millions of neurons produce a complex, irregular storm of electrical activity. Excitatory pyramidal cells transmit signals across cortical layers, fast inhibitory interneurons suppress excess firing, and subcortical inputs from the thalamus and neuromodulatory systems continuously modulate background state. Unlike a clock’s steady rhythm, cortical activity shows large trial-to-trial variability.
Despite this internal turbulence, visual perception is typically stable and coherent. We recognize faces, judge approaching vehicles, and read signs across wildly changing conditions. How the brain preserves reliable perception while remaining flexible is a central question in systems neuroscience.
If cortical dynamics were overly rigid, the system could not adapt to novel inputs. If networks fell into uncontrolled deterministic chaos, tiny variations in initial conditions would quickly amplify, producing inconsistent outputs from identical sensory inputs and undermining reliable vision.
To investigate how the brain avoids those extremes, a team led by Dr. Mehdi Borjkhani at the International Centre for Translational Eye Research (ICTER) developed a compact computational framework that tests how intrinsic feedback keeps cortical circuits close to—but not within—chaotic regimes. Their results were published in the Journal of Computational Neuroscience.
“By chaos we mean deterministic dynamics in which small initial differences rapidly diverge,” said Dr. Borjkhani. “For the brain, such dynamics can provide flexibility but also threaten stable information processing.”
A Minimal Three-Variable Circuit (E-I-M)
Rather than simulate billions of neurons, the researchers reduced primary visual cortex (V1) to three interacting variables using a modified chaotic Lotka–Volterra framework:
- E (Excitatory population): Pyramidal neurons, roughly 80% of cortical cells.
- I (Inhibitory population): Fast-spiking parvalbumin-positive (PV+) interneurons.
- M (Modulatory drive): Slower inputs combining thalamic signals and ascending neuromodulatory systems.
Across 225 parameter configurations adjusting excitation, inhibition, and baseline drive, nearly 90% produced chaotic dynamics, identified by a positive Lyapunov exponent. However, introducing two biologically motivated feedback mechanisms dramatically changed the behavior:
- Rapid excitatory-to-inhibitory recruitment: As pyramidal firing increases, local inhibitory interneurons are quickly engaged to prevent runaway excitation.
- Homeostatic regulation: Slower feedback constrains the modulatory drive so activity stays within physiological bounds.
With these mechanisms active, chaotic attractors were transformed into stable operating cycles and excitatory variance dropped from 0.325 to 0.024—a reduction of roughly 93%. Importantly, this stabilization did not require precise parameter tuning: the effect remained robust across parameter perturbations up to 25%.
“We did not remove the model’s capacity for chaotic dynamics,” Dr. Borjkhani noted. “We added the same control motifs that real cortex uses: fast inhibition and slower self-regulation.”
Validating Against Primate Visual Physiology
To confirm biological relevance, the team compared model behavior to known properties of mammalian V1. They presented 30 simulated neurons with oriented bars across ten angles (0°–180°) and computed orientation selectivity indices (OSI). The model produced average OSI values of about 0.38 ± 0.09 in the chaotic regime and 0.31 ± 0.10 after regularization—well within experimentally observed macaque V1 ranges.
Feeding the network output into a Hodgkin–Huxley spike generator, the irregular drive produced a spike irregularity coefficient (CV) of 0.27, consistent with cortical slice measurements. Thus, moderate internal irregularity coexists with accurate stimulus tuning and realistic spiking statistics.
Implications for Neural Stability and Disease
These results reframe sensory stability as an active, ongoing balance rather than a passive fixed state. Operating near the edge of instability provides cortical circuits with the computational flexibility needed to respond to rapid visual changes while preventing sensory collapse.
The model generates testable clinical hypotheses: disrupting fast interneuron feedback or slower homeostatic regulation should push cortical activity past the stability threshold and could contribute to disorders involving excitation/inhibition imbalance, such as epilepsy or certain features of psychiatric illness.
“We are not proposing an immediate treatment,” Dr. Borjkhani emphasized. “We offer a simple map of relationships and specific predictions that experiments can test. Understanding how the brain controls near-chaotic activity is as important as studying how it generates complexity.”
Editorial Notes:
- This article was edited by a Neuroscience News editor.
- The journal paper was reviewed in full by staff reviewers.
- Additional context was provided by the editorial team.
About this Genetics and Neuroregeneration Research:
- Media Contact: Marcin Bernatek
- Source: Institute of Physical Chemistry of the Polish Academy of Sciences
- Image Credit: Image credited to Neuroscience News
- Original Research (Open Access): Journal of Computational Neuroscience (September 21, 2026). “Intrinsic chaos control in cortical circuits: A minimal E-I-M rate model for primary visual cortex.” Authors: Mehdi Borjkhani, Morteza A. Sharif & Hadi Borjkhani.
- DOI: 10.1007/s10827-026-00938-5
Abstract
Intrinsic chaos control in cortical circuits: A minimal E-I-M rate model for primary visual cortex
Cortical circuits show variable yet bounded activity, implying operation near—rather than within—fully chaotic regimes. We present a minimal three-variable rate model for primary visual cortex (V1) that demonstrates how biologically plausible feedback mechanisms act as intrinsic controllers of chaos.
Starting from a simple chaotic Lotka–Volterra scaffold, we introduce three biologically grounded modifications: excitatory-to-inhibitory coupling, homeostatic regulation of modulatory drive, and orientation-tuned sensory input. These changes convert chaotic attractors for excitatory (E), inhibitory (I), and modulatory (M) populations into controlled limit cycles, yielding an approximate 93% reduction in dynamical variance.
The model reproduces essential V1 phenomena: orientation selectivity within experimental distributions, stimulus-induced reductions in variability, and realistic spike irregularity when coupled to Hodgkin–Huxley neurons. Parameter analysis shows that feedback motifs robustly stabilize activity across most of the previously chaotic regime, and that specific nonlinear disinhibition is necessary for chaos in this minimal architecture.
Overall, our findings suggest cortical circuits possess an intrinsic capacity for chaos that canonical feedback motifs actively suppress, positioning the brain at the edge of instability where flexibility and reliable signal processing coexist.