Predictable irregularities in retinal receptive fields

Liu, Yuan Sophie; Stevens, Charles F.; Sharpee, Tatyana O.

doi:10.1073/pnas.0908926106

Cited by 51 publications

(54 citation statements)

References 29 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…While OFF cells distribute more densely than their counterpart ON cells, both types have similar receptive-field overlap (spacing is about two times the SD of a Gaussian fit to the central receptive field) (6,10,11). Furthermore, OFF arbors (as we quantify here) branch more densely (5) and provide similar dendritic membrane areas as ON arbors.…”

mentioning

confidence: 88%

Retina is structured to process an excess of darkness in natural scenes

Ratliff

Borghuis

Kao

et al. 2010

Proc. Natl. Acad. Sci. U.S.A.

184

215

View full text Add to dashboard Cite

Retinal ganglion cells that respond selectively to a dark spot on a brighter background (OFF cells) have smaller dendritic fields than their ON counterparts and are more numerous. OFF cells also branch more densely, and thus collect more synapses per visual angle. That the retina devotes more resources to processing dark contrasts predicts that natural images contain more dark information. We confirm this across a range of spatial scales and trace the origin of this phenomenon to the statistical structure of natural scenes. We show that the optimal mosaics for encoding natural images are also asymmetric, with OFF elements smaller and more numerous, matching retinal structure. Finally, the concentration of synapses within a dendritic field matches the information content, suggesting a simple principle to connect a concrete fact of neuroanatomy with the abstract concept of information: equal synapses for equal bits.ganglion cells | neural coding | vision T he brain separates light from dark unequally. Psychophysical studies and measurements of visually evoked potentials show greater sensitivity to light decrements and dark spots in images (1, 2). Also, more cortical cells respond to negative than positive contrasts (3). In fact, this asymmetry begins with the second order (bipolar) neurons in the retina, which rectify the local contrast signal from the cone array (Fig. 1A). OFF cone bipolar cells (responding mostly to negative contrasts) outnumber the ON cells (responding mostly to positive contrasts) by 2-fold (4); thus, right from the start, the brain provides more resources for signaling negative contrasts.This aspect of retinal structure continues through the ganglion-cell level, where some ganglion-cell types have paired ON and OFF polarities (e.g., P and M in monkey, brisk-transient and brisk-sustained in rabbit and guinea pig, and X and Y in cat). Of these, the OFF cells have narrower dendritic fields and correspondingly narrower receptive-field centers than their ON partners [guinea pig (Fig. 1B), rat (5), rabbit (6), monkey (7), human (8), and smaller differences in cat (9)]. Thus, to cover the retina, OFF cells outnumber their ON partners. OFF arbors are narrower across cell classes with different spatiotemporal bandwidths (e.g., they are narrower for both midget and parasol cells in humans) (8). Specifically, whereas in fovea, midget cells are paired (one ON and one OFF per cone), beyond fovea, where midget cells collect from many bipolar cells and cones, OFF cells have smaller arbors and hence, are more numerous.While OFF cells distribute more densely than their counterpart ON cells, both types have similar receptive-field overlap (spacing is about two times the SD of a Gaussian fit to the central receptive field) (6, 10, 11). Furthermore, OFF arbors (as we quantify here) branch more densely (5) and provide similar dendritic membrane areas as ON arbors. Because the membrane density of excitatory synapses (synapses/μm 2 ) is constant across an arbor and across cell types (12, 13), OFF cells receive simi...

show abstract

mentioning

confidence: 88%

Retina is structured to process an excess of darkness in natural scenes

Ratliff

Borghuis

Kao

et al. 2010

Proc. Natl. Acad. Sci. U.S.A.

184

215

View full text Add to dashboard Cite

show abstract

“…Each pixel is the correlation between the interneuron and a different spatial location within a single model cell type. Because retinal cell types are not perfectly homogeneous 25,26 , the model contained for each location a parameter that scaled the receptive field amplitude. These parameters created a modest improvement in performance (0.69 ± 0.02 vs 0.66 ± 0.02 correlation coefficient without scaling parameters), and created receptive fields in layer 2 that had small differences within a cell type (Extended Data, Fig.…”

Section: Cnn Internal Units Have Receptive Field Structure Matching Rmentioning

confidence: 99%

The dynamic neural code of the retina for natural scenes

Maheswaranathan

Tanaka

et al. 2018

Preprint

View full text Add to dashboard Cite

The normal function of the retina is to convey information about natural visual images. It is this visual environment that has driven evolution, and that is clinically relevant. Yet nearly all of our understanding of the neural computations, biological function, and circuit mechanisms of the retina comes in the context of artificially structured stimuli such as flashing spots, moving bars and white noise. It is fundamentally unclear how these artificial stimuli are related to circuit processes engaged under natural stimuli. A key barrier is the lack of methods for analyzing retinal responses to natural images. We addressed both these issues by applying convolutional neural network models (CNNs) to capture retinal responses to natural scenes. We find that CNN models predict natural scene responses with high accuracy, achieving performance close to the fundamental limits of predictability set by intrinsic cellular variability. Furthermore, individual internal units of the model are highly correlated with actual retinal interneuron responses that were recorded separately and never presented to the model during training. Finally, we find that models fit only to natural scenes, but not white noise, reproduce a range of phenomena previously described using distinct artificial stimuli, including frequency doubling, latency encoding, motion anticipation, fast contrast adaptation, synchronized responses to motion reversal and object motion sensitivity. Further examination of the model revealed extremely rapid context dependence of retinal feature sensitivity under natural scenes using an analysis not feasible from direct examination of retinal responses. Overall, these results show that nonlinear retinal processes engaged by artificial stimuli are also engaged in and relevant to natural visual processing, and that CNN models form a powerful and unifying tool to study how sensory circuitry produces computations in a natural context.

show abstract

“…To achieve this, interactions in a network should be organized to exploit existing correlations in neural inputs to compensate for noise-induced errors. Such a trade-off between decorrelation and noise reduction possibly accounts for the organization of several biological information processing systems, e.g., the adaptation of center-surround receptive fields to ambient light intensity (12)(13)(14), the structure of retinal ganglion cell mosaics (15)(16)(17)(18), and the genetic regulatory network in a developing fruit fly (19,20). In engineered systems, compression (to decorrelate incoming data stream), followed by reintroduction of error-correcting redundancy, is an established way of building efficient codes (21).…”

mentioning

confidence: 99%

Optimal population coding by noisy spiking neurons

Tkačik

Prentice

Balasubramanian

et al. 2010

Proc. Natl. Acad. Sci. U.S.A.

160

176

View full text Add to dashboard Cite

In retina and in cortical slice the collective response of spiking neural populations is well described by "maximum-entropy" models in which only pairs of neurons interact. We asked, how should such interactions be organized to maximize the amount of information represented in population responses? To this end, we extended the linear-nonlinear-Poisson model of single neural response to include pairwise interactions, yielding a stimulus-dependent, pairwise maximum-entropy model. We found that as we varied the noise level in single neurons and the distribution of network inputs, the optimal pairwise interactions smoothly interpolated to achieve network functions that are usually regarded as discrete-stimulus decorrelation, error correction, and independent encoding. These functions reflected a trade-off between efficient consumption of finite neural bandwidth and the use of redundancy to mitigate noise. Spontaneous activity in the optimal network reflected stimulus-induced activity patterns, and single-neuron response variability overestimated network noise. Our analysis suggests that rather than having a single coding principle hardwired in their architecture, networks in the brain should adapt their function to changing noise and stimulus correlations.adaptation | neural networks | Ising model | attractor states P opulations of sensory neurons encode information about stimuli into sequences of action potentials, or spikes (1). Experiments with pairs or small groups of neurons have observed many different coding strategies (2-6): (i) independence, where each neuron responds independently to the stimulus, (ii) decorrelation, where neurons interact to give a decorrelated representation of the stimulus, (iii) error correction, where neurons respond redundantly, in patterns, to combat noise, and (iv) synergistic coding, where population activity patterns carry information unavailable from separate neurons.How should a network arrange its interactions to best represent an ensemble of stimuli? Theoretically, there has been controversy over what is the "correct" design principle for neural population codes (7-11). On the one hand, neurons have a limited repertoire of response patterns, and information is maximized by using each neuron to represent a different aspect of the stimulus. To achieve this, interactions in a network should be organized to remove correlations in network inputs and thus create a decorrelated network response. On the other hand, neurons are noisy, and noise is combatted via redundancy, where different patterns related by noise encode the same stimulus. To achieve this, interactions in a network should be organized to exploit existing correlations in neural inputs to compensate for noise-induced errors. Such a trade-off between decorrelation and noise reduction possibly accounts for the organization of several biological information processing systems, e.g., the adaptation of center-surround receptive fields to ambient light intensity (12-14), the structure of retinal ganglion cell mosaics (15-18),...

show abstract

Predictable irregularities in retinal receptive fields

Cited by 51 publications

References 29 publications

Retina is structured to process an excess of darkness in natural scenes

Retina is structured to process an excess of darkness in natural scenes

The dynamic neural code of the retina for natural scenes

Optimal population coding by noisy spiking neurons

Contact Info

Product

Resources

About