# Receptive fields from single-neuron recording and MRI reveal similar information coding for binocular depth

## TL;DR

This study shows that MRI-based receptive fields in humans and single-neuron recordings in macaques reveal similar patterns for processing binocular depth.

## Contribution

The study extends population receptive field (pRF) methods to binocular depth and validates them against electrophysiological data.

## Key findings

- Human and macaque V1 showed similar encoding of binocular depth.
- pRFs in higher visual areas showed greater responsiveness to relative depth.
- pRFs were more sensitive to fine-scale depth differences than electrophysiological measures.

## Abstract

The concept of a receptive field (RF) of a neuron is fundamental in sensory neuroscience, defined by the ordered set of sensory stimuli that reliably induce activation of a neuron. Recent advances in magnetic resonance (MR) imaging have developed methods for isolating cortical activations into RF-like entities. Using binocular vision, we extend the mapping of these MR-based RFs into the third dimension of stereoscopic depth and demonstrate substantial alignments of the resulting response profiles of MR-based RFs with earlier electrophysiological measures of neuronal RFs. The comparison between noninvasive measurements in the human cortex and single-neuron recordings in macaque is an essential step toward validation of the newer methods.

The population receptive field (pRF) approach to functional measurement of the sensory properties of magnetic resonance (MR)-identified locations in the human brain was extended to include the third dimension of binocular depth. In total, pRFs were extracted from nine different visual areas (V1, V2, V3, V3AB, V4, V5, V7, Ventral Occipital Cortex: VOC, Lateral Occipital Cortex: LOC) of the human cortex and, where possible, comparisons were made with electrophysiological recordings from homologous areas in the macaque cortex. Human and macaque V1 showed strikingly similar information profiles for the encoding of binocular depth. Further, both human and macaque V5 showed consistent changes in preferred binocular depth of the stimulus, dependent on whether the stimuli were binocularly correlated or anticorrelated. Across the nine areas of the visual cortex explored, the population profiles of pRFs for binocular depth showed evidence of a greater responsiveness to relative depth in higher visual cortical areas, again consistent with the findings from macaque electrophysiology. Overall, the pRF measures of cortical response were more sensitive to fine-scale differences of binocular depth, compared with many existing electrophysiological measures of tuning for binocular depth. Our results show that the pRF method can be extended beyond the characterization of RFs in retinotopic coordinates to reveal higher-order, derived visual properties. The parallels between noninvasive, MR-based measures of pRFs in humans and the electrophysiological recordings of single neurons in experimental animals make a further step toward validation of the pRF methodology.