Alert Response to Motion Onset in the Retina

Abstract

Previous studies have shown that motion onset is very effective at capturing attention and is more salient than smooth motion. Here, we find that this salience ranking is present already in the firing rate of retinal ganglion cells. By stimulating the retina with a bar that appears, stays still, and then starts moving, we demonstrate that a subset of salamander retinal ganglion cells, fast OFF cells, responds significantly more strongly to motion onset than to smooth motion. We refer to this phenomenon as an alert response to motion onset. We develop a computational model that predicts the time-varying firing rate of ganglion cells responding to the appearance, onset, and smooth motion of a bar. This model, termed the Adaptive Cascade Model (ACM), consists of a ganglion cell that receives input from a layer of bipolar cells, represented by individual rectified subunits. Additionally, both the bipolar and ganglion cells have separate contrast gain control mechanisms. This model captured the responses to our different motion stimuli over a wide range of contrasts, speeds, and locations. The alert response to motion onset, together with its computational model, introduces a new mechanism of sophisticated motion processing that occurs early in the visual system.

General features of the retinal connectome determine the computation of motion anticipation

“…We do not understand how the retina generates motion anticipation, but the ubiquity of this process across ganglion cell types suggests that it reflects general properties of the inner retinal circuit rather than the specific wiring of subtypes of neuron. A fast decrease in the gain with which signals are transmitted through the retina has been proposed to account for motion anticipation, but the site(s) of such control have not been identified ( Berry et al, 1999 ; Schwartz et al, 2007 ; Chen et al, 2013 ). One possibility is that moving stimuli induce a fast decrease in the efficiency of excitatory transmission from bipolar cells to ganglion cells, shortening the time-course of excitation.…”

“…This idea is attractive because bipolar cells are the only route by which excitatory signals are transmitted to ganglion cells, and these synapses have been identified as a major site of gain-control, with an increase in temporal contrast causing fast depression of vesicle release ( Demb, 2008 ; Jarsky et al, 2011 ; Nikolaev et al, 2013 ). A second possibility is that some mechanism intrinsic to ganglion cells alters the time-course of the ganglion cell response ( Chen et al, 2013 ). We now need to test these various possibilities experimentally.…”

Clustering of Neural Activity: A Design Principle for Population Codes

“…To the extent that such models accurately capture the responses of many neurons over a broad set of stimuli, then this version also constitutes a solution to the neural code. While there are notable examples of cell types and stimulus sets for which such models are highly successful, it is unclear how well this approach will work for the full set of behaviorally relevant stimuli and whether it correctly captures the correlation structure present in the neural population (Rodieck, 1965;van Hateren et al, 2002;Pillow et al, 2008;Chen et al, 2013;McIntosh et al, 2016;Buckley and Toyoizumi, 2018). But regardless of the success of encoding models, this approach is not well matched to the way that the brain works.…”

Nonlinear transfer and temporal gain control in ON bipolar cells

“…Their axon terminals can convert membrane voltage into a graded release of glutamate in a manner embodying rectification of the visual signal. As a result of this rectification, bipolar cells act as nonlinear subunits within a ganglion cell's receptive field [2][3][4][5], thereby giving rise to forms of translation invariance in retinal motion processing [6][7][8]. Furthermore, synaptic depression within the axon terminal can dramatically reduce glutamate release [9][10][11], giving rise to a form of gain control that emphasizes motion discontinuities [8,12].…”