Population receptive field (pRF) measurements of chromatic responses in human visual cortex using fMRI

Lauren E Welbourne,Antony B Morland,Alex R Wade

doi:10.1016/j.neuroimage.2017.11.022

Lauren E Welbourne, Antony B Morland + Show 1 more

Open Access

https://doi.org/10.1016/j.neuroimage.2017.11.022

Copy DOI

Abstract

The spatial sensitivity of the human visual system depends on stimulus color: achromatic gratings can be resolved at relatively high spatial frequencies while sensitivity to isoluminant color contrast tends to be more low-pass. Models of early spatial vision often assume that the receptive field size of pattern-sensitive neurons is correlated with their spatial frequency sensitivity - larger receptive fields are typically associated with lower optimal spatial frequency. A strong prediction of this model is that neurons coding isoluminant chromatic patterns should have, on average, a larger receptive field size than neurons sensitive to achromatic patterns. Here, we test this assumption using functional magnetic resonance imaging (fMRI). We show that while spatial frequency sensitivity depends on chromaticity in the manner predicted by behavioral measurements, population receptive field (pRF) size measurements show no such dependency. At any given eccentricity, the mean pRF size for neuronal populations driven by luminance, opponent red/green and S-cone isolating contrast, are identical. Changes in pRF size (for example, an increase with eccentricity and visual area hierarchy) are also identical across the three chromatic conditions. These results suggest that fMRI measurements of receptive field size and spatial resolution can be decoupled under some circumstances - potentially reflecting a fundamental dissociation between these parameters at the level of neuronal populations.

Highlights

The three pathways that contribute to human color vision originate in different retinal combinations of the signals from the long, medium and short-wave sensitive cone photoreceptors (Hurvich and Jameson, 1957; Jameson and Hurvich, 1968)
Condition, and spatial frequency were entered into a repeated-measures ANOVA to identify the effect of each on the contrast sensitivity values; significant main effects were found for each factor (eccentricity (F(1,4) 1⁄4 179.921, p 1⁄4 10À4), condition (F(2,8) 1⁄4 78.730, p 1⁄4 10À6), and spatial frequency (F(2,8) 1⁄4 153.965, p 1⁄4 10À6))
They serve as a useful validation of our stimulus generation and presentation pathway: significant errors in, for example, our calibration procedures would have led to luminance contamination of our nominally-isoluminant stimuli and a corresponding increase in similarity between the luminance and isoluminant spatial sensitivity functions

Summary

Introduction

The three pathways that contribute to human color vision originate in different retinal combinations of the signals from the long, medium and short-wave sensitive cone photoreceptors (Hurvich and Jameson, 1957; Jameson and Hurvich, 1968). One pathway processes achromatic luminance (L þ M), and two are isoluminant chromatic pathways: ‘red vs green’ (L-M) and ‘yellow vs blue’ (S-cone isolating) These precortical physiological pathways can be probed by psychophysical experiments, which demonstrate differences in their spatial frequency tuning profiles (Johnson et al, 2010; Kim et al, 2013; Mullen, 1985; Owsley et al, 1983; Poirson and Wandell, 1993, 1996; Webster et al, 1990). This non-linearity is typical of complex cells, in which spatial tuning is independent of receptive field sizes (Movshon et al, 1978)

Methods

Results

Conclusion