Spatial receptive field structure of double-opponent cells in macaque V1

Abstract

Double-opponent (DO) cells are well-suited to contribute to the spatial processing of color due to their spatially opponent and cone-opponent receptive fields (RFs).However, the representation of visual images by DO cells in monkey primary visual cortex is unclear because the spatial RF structure of DO cells has not been fully characterized. Early reports suggested that DO cells have center-surround RFs, but more recent studies have shown that some DO cells are orientation-tuned. To characterize the RFs of DO cells, we mapped them in awake fixating macaques, fit them with parametric models, and compared them to the benchmark of orientationtuned simple cells. Neurons were stimulated with colorful, dynamic white noise patterns.Spike-triggered averaging was used to classify each neuron as simple, DO, or neither and to measure its spatial RF. The spatial RF of each neuron was fitted with a Gabor model and a Difference of Gaussians (DoG) model. The Gabor model provided the more accurate description, a result that is incompatible with a center-surround RF organization. The superiority of the Gabor fits was slightly more decisive for simple than for DO cells. A modified (non-concentric) DoG model performed nearly as well as the Gabor model for DO cells.

Coding of chromatic spatial contrast by macaque V1 neurons

“…Whether real V1 neurons respond in accordance with the energy model or show enhanced responses to natural alignments between chromatic and luminance signals is an important and unanswered question.In summary, DO cells and simple cells have much in common. As shown by the current study, both cell types combine signals roughly linearly across their RFs and, as shown previously, they share a Gabor-like RF structure(De and Horwitz, 2020). These observations motivate the idea that simple cells and DO cells are closely related neuronal types that may contribute similarly to downstream circuits that integrate coneopponent and cone-non-opponent signals for spatial image analysis.…”

“…By definition, simple cells have large, non-opponent L- and M-cone weights, and DO cells have cone-opponent weights. We used the same cone weight criteria that we used previously to facilitate comparison between studies (De and Horwitz, 2020).…”

“…Despite its importance for visual function, the spatial integration of spectral signals by neurons in the visual system is poorly understood. (Daw, 1968;Conway, 2001; De and Horwitz, 2020).…”

“…The first stage of the primate visual system at which spatial comparison of spectral signals is implemented by individual neurons is area V1; many neurons in V1 are tuned for orientation and spectral properties of edges. Simple and double-opponent (DO) cells are tuned for the orientation of luminance and chromatic edges, respectively (Ringach, 2002; Conway and Livingstone, 2006; Johnson et al, 2008; De and Horwitz, 2020). Simple cells combine signals from the L- and M-cone photoreceptors with the same sign and are therefore cone non-opponent.…”

“…They respond to light increments in one part of their receptive fields (RFs) and decrements in another, rendering them sensitive to spatial luminance contrast. DO cells are spatially opponent, like simple cells are, but they combine signals from at least two different cone photoreceptor types antagonistically within individual RF subfields (Daw, 1968; Conway, 2001; De and Horwitz, 2020).…”