Normal binocular rivalry in autism: implications for the excitation/inhibition imbalance hypothesis

Abstract

Autism is characterized by disruption in multiple dimensions of perception, emotion, language and social cognition. Many hypotheses for the underlying neurophysiological basis have been proposed. Among these is the excitation/inhibition (E/I) imbalance hypothesis, which states that levels of cortical excitation and inhibition are disrupted in autism. We tested this theory in the visual system, because vision is one of the better understood systems in neuroscience, and because the E/I imbalance theory has been proposed to explain hypersensitivity to sensory stimuli in autism. We conducted two experiments on binocular rivalry, a well-studied psychophysical phenomenon that depends critically on excitation and inhibition levels in cortex. Using a computational model, we made specific predictions about how imbalances in excitation and inhibition levels would affect perception during two aspects of binocular rivalry: mixed perception (Experiment 1) and traveling waves (Experiment 2). We found no significant differences in either of these phenomena between high-functioning adults with autism and controls, and no evidence for a relationship between these measurements and the severity of autism. These results do not conclusively rule out an excitation/inhibition imbalance in the visual system of those with autism, but they suggest that such an imbalance, if it exists, is likely to be small in magnitude.

Figure 1

Experiment 1 stimulus. Subjects wore prism goggles so that the left grating was presented to the left eye and the right grating was presented to the right eye at corresponding locations in the two eyes. At any given time, subjects perceived either the grating tilted clockwise of vertical, counterclockwise of vertical, or a mixture of the two (typically perceived as a plaid).

Figure 2

Experiment 2 stimuli. (A) Green spiral grating, shown here to the left eye. Red radial grating, shown to the right eye. The green spiral percept was initially dominant on most trials. (B). Brief contrast increment at the top of the right eye stimulus triggered a perceptual traveling wave. (C) Example of traveling waves percept following the contrast increment. The traveling wave was perceived even though the stimulus was physically unchanging. Subject pressed a button when the wave crossed the tick mark (bottom left in this example). The tick mark was at different locations on different trials to estimate wave speed.

Figure 3

Model and model predictions. (A) Schematic of computational model. Dark gray neurons respond to one stimulus orientation and light gray neurons respond to the orthogonal orientation. While this schematic shows an 8-neuron model for simplicity, mixed perception was simulated with a 20-neuron model, and traveling waves were simulated with a 400-neuron model. (B) Effects of excitatory and inhibitory weight strengths on model mixed perception. Lighter shading indicates a higher percentage of time in a mixed percept state. (C) Effects of excitatory and inhibitory weight strengths on traveling wave speed. Lighter shading indicates faster waves.

Figure 4

(A) Experiment 1: Percentage of the time subjects perceived a mixture of the two stimuli. (B) Experiment 2: Subject response latency as a function of tick mark position. Blue, autism group. Brown, control group. Shaded regions represent standard error of the mean. Wave speed was estimated by computing the inverse slope of this function. (C) Experiment 2: Wave speed, averaged across subjects. Error bars, standard error of the mean.