Altered functional interactions between neurons in primary visual cortex of macaque monkeys with experimental amblyopia

Katerina Acar, Lynne Kiorpes, J. Anthony Movshon, Matthew A. Smith

Abstract

Amblyopia is a developmental disorder characterized by degraded vision through one eye, typically resulting from early-life strabismus (misalignment) or anisometropia. While it is well-established that single-neuron responses in the primary visual cortex (V1) are reduced when driven by the amblyopic eye, these individual deficits do not fully account for the significant losses in visual sensitivity and resolution observed behaviorally. This study investigates whether altered patterns of coordinated activity across neuronal populations contribute to these deficits. By recording from V1 in macaque monkeys with experimental strabismic amblyopia, we found that activity driven by the amblyopic eye exhibits higher trial-to-trial variability and increased interneuronal correlations compared to the fellow eye or healthy controls. These changes in functional interactions significantly diminish the visual information available to downstream decoders, suggesting a systemic population-level basis for amblyopic vision loss.

Introduction

Amblyopia is often described as a "lazy eye," but the pathology resides within the visual cortex rather than the eye itself. In macaque models of strabismic amblyopia, V1 neurons driven by the affected eye show reduced responsivity and a loss of binocularity. However, a persistent paradox remains: the degradation of single-unit firing rates is often too mild to explain the severe behavioral impairments in contrast sensitivity and acuity.

This research explores the hypothesis that the neural deficit in amblyopia is not merely a matter of "weak signals" from individual cells, but a failure of population dynamics. If neurons driven by the amblyopic eye fire with excessive redundancy or high noise correlations, the brain's ability to decode visual information would be severely compromised. We examined the functional interactions—specifically noise correlations and trial-to-trial variability—within V1 populations to determine how these dynamics impact the representation of visual stimuli.

Methods

Subjects and Electrophysiology

The study utilized four adult macaque monkeys (M. nemestrina): three with surgically induced strabismus (experimental subjects EM 640, JS 579, and HN 580) and one healthy control (LU 645). Researchers implanted 100-electrode "Utah" arrays into the superficial layers of V1 to record simultaneous activity from neuronal populations.

Visual Stimulation

Monkeys were presented with drifting sinusoidal gratings at 12 different orientations. To assess the impact of signal strength on population dynamics, stimuli were presented at three contrast levels: 12%, 50%, and 100%.

Statistical and Decoding Analysis

Two primary metrics were used to quantify population noise:

  1. Fano Factor (FF): A measure of trial-to-trial variability in individual neurons.
  2. Noise Correlation ($r_{sc}$): A measure of the degree to which pairs of neurons fluctuate together in their response to identical stimuli.

To evaluate the functional impact of these metrics, a Support Vector Machine (SVM) classifier with Error-Correcting Output Coding (ECOC) was employed to decode stimulus orientation from the activity of 30-neuron ensembles.

Results

Reduced Responsivity and Increased Variability

Consistent with previous literature, mean maximum firing rates were significantly lower for the amblyopic eye (9.56 sp/s) compared to the fellow eye (15.08 sp/s). Furthermore, the amblyopic eye exhibited a higher Fano Factor, indicating greater trial-to-trial instability. Crucially, this increased variability persisted even after using mean-matching techniques to control for differences in firing rates, suggesting that the noise is a fundamental property of the amblyopic circuit rather than a simple byproduct of lower activity.

Altered Noise Correlations

In healthy visual systems, increasing stimulus contrast typically "quenches" noise, reducing correlations between neurons and sharpening the population signal. In the amblyopic eye, however, noise correlations ($r_{sc}$) were significantly higher (0.21) than in the fellow eye (0.16).

Perhaps more importantly, the amblyopic eye's correlations were remarkably insensitive to contrast. While the fellow eye showed the expected decrease in $r_{sc}$ as contrast increased, the correlation-contrast curve for the amblyopic eye remained nearly flat. This suggests a "rigid" or "stiff" network state that fails to engage properly even when presented with high-quality visual input.

Information Loss

The decoding analysis revealed that these population-level changes have a direct impact on visual representation. Orientation classification accuracy was consistently lower for the amblyopic eye. For example, in representative sessions, accuracy dropped from approximately 85% when driven by the fellow eye to 65% for the amblyopic eye. The high degree of correlation among neurons with similar orientation tuning in the amblyopic eye created redundancy; the neurons were essentially "shouting the same thing," including the same noise, making it harder for downstream areas to extract a clean signal.

Discussion

These findings suggest that the neural basis of amblyopia involves a "fifth factor" beyond the traditionally recognized deficits of reduced responsivity, loss of binocularity, undersampling, and topographical jitter. This fifth factor—altered functional interactions—represents a systemic failure in how V1 populations coordinate their activity.

The persistence of high noise correlations and variability in the amblyopic eye suggests that the cortical network is in a state of chronic instability. Because the network fails to suppress noise in response to strong visual stimuli, the information relayed to higher visual areas is fundamentally degraded. This population-level perspective provides a more comprehensive explanation for the behavioral deficits observed in amblyopia and highlights the importance of considering circuit dynamics, rather than just single-cell metrics, in the study of developmental visual disorders. Future research may investigate whether therapeutic interventions, such as perceptual learning, can successfully "normalize" these correlation patterns to improve visual function.

Summary updated

Abstract

Abstract Amblyopia, a disorder in which vision through one of the eyes is degraded, arises because of defective processing of information by the visual system. Amblyopia often develops in humans after early misalignment of the eyes (strabismus), and can be simulated in macaque monkeys by artificially inducing strabismus. In such amblyopic animals, single-unit responses in primary visual cortex (V1) are appreciably reduced when evoked by the amblyopic eye compared to the other (fellow) eye. However, this degradation in single V1 neuron responsivity is not commensurate with the marked losses in visual sensitivity and resolution measured behaviorally. Here we explored the idea that changes in patterns of coordinated activity across populations of V1 neurons may contribute to degraded visual representations in amblyopia, potentially making it more difficult to read out evoked activity to support perceptual decisions. We studied the visually-evoked activity of V1 neuronal populations in three macaques ( M. nemestrina ) with strabismic amblyopia and in one control. Activity driven through the amblyopic eye was diminished, and these responses also showed more interneuronal correlation at all stimulus contrasts than responses driven through the fellow eye or responses in the control. A decoding analysis showed that responses driven through the amblyopic eye carried less visual information than other responses. Our results suggest that part of the reduced visual capacity of amblyopes may be due to changes in the patterns of functional interaction among neurons in V1. New and noteworthy Amblyopia is a developmental disorder of visual processing that reduces visual function and changes the visual responses of cortical neurons in macaque monkeys. The neuronal and behavioral changes are not always well correlated. We found that the interactions among neurons in the visual cortex of monkeys with amblyopia are also altered. These changes may contribute to amblyopic visual deficits by diminishing the amount of information relayed by neuronal populations driven by the amblyopic eye.
BibTeX
@article{acar2019altered,
  author = {Katerina Acar and Lynne Kiorpes and J. Anthony Movshon and Matthew A. Smith},
  title = {{Altered functional interactions between neurons in primary visual cortex of macaque monkeys with experimental amblyopia}},
  journal = {Journal of Neurophysiology},
  year = {2019},
  doi = {10.1152/jn.00232.2019},
}
Figure