Abstract
The perceptual salience and visibility of image elements is influenced by other elements in their vicinity. While collinear flanking elements usually improve sensitivity for a target in foveal vision, this study finds that in the parafovea, these flankers primarily exert suppressive effects. Recording from V1 neurons in macaques, the researchers demonstrate that collinear flankers reduce perceptual sensitivity by increasing baseline neural activity, which in turn reduces the dynamic range available for responding to the central target.
Introduction
Our visual system does not process objects in isolation; the context surrounding an element significantly alters its perception. A well-documented phenomenon in foveal psychophysics is facilitation: a faint target becomes easier to detect when flanked by collinear elements. However, a discrepancy exists between these human behavioral findings and animal neurophysiology, which often reports suppression at more eccentric parafoveal locations. This study investigates this gap by recording from the primary visual cortex (V1) of macaques during contrast detection tasks, specifically examining why the parafovea behaves differently than the fovea.
Methods
The researchers employed a dual approach using rhesus macaques. Single-unit recordings were taken from 72 V1 neurons during passive fixation to map Contrast Response Functions (CRFs). Additionally, multiunit recordings were obtained from 38 V1 sites while monkeys performed a contrast detection task, reporting the presence of a faint Gabor target via eye movements. Stimuli consisted of a central Gabor target and two collinear flankers placed at varying distances (1, 2, or 3 wavelengths, $\lambda$).
Results
The study revealed that parafoveal V1 is dominated by lateral inhibition. Even at distances typically associated with facilitation in the fovea (3$\lambda$), flankers induced suppression.
A key finding was the mechanism of this suppression: flankers increased "ongoing activity" in the absence of a target (raising the "floor"), but simultaneously reduced the neuron's maximum firing capacity when the target appeared (lowering the "ceiling"). This reduction in dynamic range directly correlated with decreased behavioral sensitivity. Furthermore, the increased baseline activity acted as neural "noise," leading to a higher rate of false alarms in the detection task. The suppressive effect was strongest when flankers matched the neuron's preferred orientation (collinear) and weakest when they were orthogonal.
Discussion
These results suggest that the dominant effect of collinear flankers in parafoveal vision is suppression, contrasting with the facilitation observed at the fovea. This difference may stem from distinct neural "wiring" or the influence of top-down attention, which might override suppressive signals in the center of gaze. Future research is needed to directly compare foveal and parafoveal V1 using identical recording techniques to resolve whether these differences are inherent to the cortical architecture or modulated by behavioral state.
Summary updated