A synaptic explanation of suppression in visual cortex

Response sub-additivity and variability quenching in visual cortex

Sub-additivity and variability are ubiquitous response motifs in the primary visual cortex (V1). Response sub-additivity enables the construction of useful interpretations of the visual environment, whereas response variability indicates the factors that limit the precision with which the brain can do this. There is increasing evidence that experimental manipulations that elicit response sub-additivity often also quench response variability. Here, we provide an overview of these phenomena and suggest that they may have common origins. We discuss empirical findings and recent model-based insights into the functional operations, computational objectives and circuit mechanisms underlying V1 activity. These different modelling approaches all predict that response sub-additivity and variability quenching often co-occur. The phenomenology of these two response motifs, as well as many of the insights obtained about them in V1, generalize to other cortical areas. Thus, the connection between response sub-additivity and variability quenching may be a canonical motif across the cortex.

Investigating cross-orientation inhibition with continuous tracking

We investigated cross-orientation inhibition with the recently developed continuous tracking technique. We designed an experiment where participants tracked the horizontal motion of a narrow vertical grating. The target was superimposed on one of three different backgrounds, in separate sessions: a uniform gray background or a sinusoidal grating oriented either parallel or orthogonal to the target. Both mask and target where phase reversed. We cross-correlated target and mouse movements and compared the peaks and lags of response with the different masks. Our results are in agreement with previous findings on cross-orientation inhibition: The orthogonal mask had a weak effect on the peaks and lags of correlation as a function of target contrast, consistently with a divisive effect of the mask, while the parallel mask acted subtractively on the response. Interestingly, lags of correlation decreased approximately linearly with contrast, with decrements of the order of 100 ms, even at 10 times the detection threshold, confirming that it is possible to investigate behavioral differences above threshold using the continuous tracking paradigm.

A power law describes the magnitude of adaptation in neural populations of primary visual cortex

How do neural populations adapt to the time-varying statistics of sensory input? We used two-photon imaging to measure the activity of neurons in mouse primary visual cortex adapted to different sensory environments, each defined by a distinct probability distribution over a stimulus set. We find that two properties of adaptation capture how the population response to a given stimulus, viewed as a vector, changes across environments. First, the ratio between the response magnitudes is a power law of the ratio between the stimulus probabilities. Second, the response direction to a stimulus is largely invariant. These rules could be used to predict how cortical populations adapt to novel, sensory environments. Finally, we show how the power law enables the cortex to preferentially signal unexpected stimuli and to adjust the metabolic cost of its sensory representation to the entropy of the environment.

Input-specific synaptic depression shapes temporal integration in mouse visual cortex

Efficient sensory processing requires the nervous system to adjust to ongoing features of the environment. In primary visual cortex (V1), neuronal activity strongly depends on recent stimulus history. Existing models can explain effects of prolonged stimulus presentation but remain insufficient for explaining effects observed after shorter durations commonly encountered under natural conditions. We investigated the mechanisms driving adaptation in response to brief (100 ms) stimuli in L2/3 V1 neurons by performing in vivo whole-cell recordings to measure membrane potential and synaptic inputs. We find that rapid adaptation is generated by stimulus-specific suppression of excitatory and inhibitory synaptic inputs. Targeted optogenetic experiments reveal that these synaptic effects are due to input-specific short-term depression of transmission between layers 4 and 2/3. Thus, brief stimulus presentation engages a distinct adaptation mechanism from that previously reported in response to prolonged stimuli, enabling flexible control of sensory encoding across a wide range of timescales.

A power law of cortical adaptation

How do neural populations adapt to the time-varying statistics of sensory input? To investigate, we measured the activity of neurons in primary visual cortex adapted to different environments, each associated with a distinct probability distribution over a stimulus set. Within each environment, a stimulus sequence was generated by independently sampling form its distribution. We find that two properties of adaptation capture how the population responses to a given stimulus, viewed as vectors, are linked across environments. First, the ratio between the response magnitudes is a power law of the ratio between the stimulus probabilities. Second, the response directions are largely invariant. These rules can be used to predict how cortical populations adapt to novel, sensory environments. Finally, we show how the power law enables the cortex to preferentially signal unexpected stimuli and to adjust the metabolic cost of its sensory representation to the entropy of the environment.

Input-specific synaptic depression shapes temporal integration in mouse visual cortex

Efficient sensory processing requires the nervous system to adjust to ongoing features of the environment. In primary visual cortex (V1), neuronal activity strongly depends on recent stimulus history. Existing models can explain effects of prolonged stimulus presentation, but remain insufficient for explaining effects observed after shorter durations commonly encountered under natural conditions. We investigated the mechanisms driving adaptation in response to brief (100 ms) stimuli in L2/3 V1 neurons by performing in vivo whole-cell recordings to measure membrane potential and synaptic inputs. We find that rapid adaptation is generated by stimulus-specific suppression of excitatory and inhibitory synaptic inputs. Targeted optogenetic experiments reveal that these synaptic effects are due to input-specific short-term depression of transmission between layers 4 and 2/3. Thus, distinct mechanisms are engaged following brief and prolonged stimulus presentation and together enable flexible control of sensory encoding across a wide range of time scales.

Deeper neural network models better reflect how humans cope with contrast variation in object recognition

Visual inputs are far from ideal in everyday situations such as in the fog where the contrasts of input stimuli are low. However, human perception remains relatively robust to contrast variations. To provide insights about the underlying mechanisms of contrast invariance, we addressed two questions. Do contrast effects disappear along the visual hierarchy? Do later stages of the visual hierarchy contribute to contrast invariance? We ran a behavioral experiment where we manipulated the level of stimulus contrast and the involvement of higher-level visual areas through immediate and delayed backward masking of the stimulus. Backward masking led to significant drop in performance in our visual categorization task, supporting the role of higher-level visual areas in contrast invariance. To obtain mechanistic insights, we ran the same categorization task on three state-of the-art computational models of human vision each with a different depth in visual hierarchy. We found contrast effects all along the visual hierarchy, no matter how far into the hierarchy. Moreover, that final layers of deeper hierarchical models, which had been shown to be best models of final stages of the visual system, coped with contrast effects more effectively. These results suggest that, while contrast effects reach the final stages of the hierarchy, those stages play a significant role in compensating for contrast variations in the visual system.

Neural substrates of perception in the vestibular thalamus during natural self-motion: A review

Accumulating evidence across multiple sensory modalities suggests that the thalamus does not simply relay information from the periphery to the cortex. Here we review recent findings showing that vestibular neurons within the ventral posteriolateral area of the thalamus perform nonlinear transformations on their afferent input that determine our subjective awareness of motion. Specifically, these neurons provide a substrate for previous psychophysical observations that perceptual discrimination thresholds are much better than predictions from Weber's law. This is because neural discrimination thresholds, which are determined from both variability and sensitivity, initially increase but then saturate with increasing stimulus amplitude, thereby matching the previously observed dependency of perceptual self-motion discrimination thresholds. Moreover, neural response dynamics give rise to unambiguous and optimized encoding of natural but not artificial stimuli. Finally, vestibular thalamic neurons selectively encode passively applied motion when occurring concurrently with voluntary (i.e., active) movements. Taken together, these results show that the vestibular thalamus plays an essential role towards generating motion perception as well as shaping our vestibular sense of agency that is not simply inherited from afferent input.

Effect of whole-hand water flow stimulation on the neural balance between excitation and inhibition in the primary somatosensory cortex

Sustained peripheral somatosensory stimulations, such as high-frequency repetitive somatosensory stimulation (HF-RSS) and vibrated stimulation, are effective in altering the balance between excitation and inhibition in the somatosensory cortex (S1) and motor cortex (M1). A recent study reported that whole-hand water flow (WF) stimulation induced neural disinhibition in the M1. Based on previous results, we hypothesized that whole-hand WF stimulation would lead to neural disinhibition in the S1 because there is a strong neural connection between M1 and S1 and aimed to examine whether whole-hand WF stimulation would change the neural balance between excitation and inhibition in the S1. Nineteen healthy volunteers were studied by measuring excitation and inhibition in the S1 before and after each of the four 15-min interventions. The excitation and inhibition in the S1 were assessed using somatosensory evoked potentials (SEPs) and paired-pulse inhibition (PPI) induced by single- and paired-pulse stimulations, respectively. The four interventions were as follows: control, whole-hand water immersion, whole-hand WF, and HF-RSS. The results showed no significant changes in SEPs and PPI following any intervention. However, changes in PPI with an interstimulus interval (ISI) of 30 ms were significantly correlated with the baseline value before whole-hand WF. Thus, the present findings indicated that the whole-hand WF stimulation had a greater decreased neural inhibition in participants with higher neural inhibition in the S1 at baseline. Considering previous results on M1, the present results possibly show that S1 has lower plasticity than M1 and that the duration (15 min) of each intervention may not have been enough to alter the balance of excitation and inhibition in the S1.

Sensory adaptation mediates efficient and unambiguous encoding of natural stimuli by vestibular thalamocortical pathways

Sensory systems must continuously adapt to optimally encode stimuli encountered within the natural environment. The prevailing view is that such optimal coding comes at the cost of increased ambiguity, yet to date, prior studies have focused on artificial stimuli. Accordingly, here we investigated whether such a trade-off between optimality and ambiguity exists in the encoding of natural stimuli in the vestibular system. We recorded vestibular nuclei and their target vestibular thalamocortical neurons during naturalistic and artificial self-motion stimulation. Surprisingly, we found no trade-off between optimality and ambiguity. Using computational methods, we demonstrate that thalamocortical neural adaptation in the form of contrast gain control actually reduces coding ambiguity without compromising the optimality of coding under naturalistic but not artificial stimulation. Thus, taken together, our results challenge the common wisdom that adaptation leads to ambiguity and instead suggest an essential role in underlying unambiguous optimized encoding of natural stimuli.

Hearing in Complex Environments: Auditory Gain Control, Attention, and Hearing Loss

Listening in noisy or complex sound environments is difficult for individuals with normal hearing and can be a debilitating impairment for those with hearing loss. Extracting meaningful information from a complex acoustic environment requires the ability to accurately encode specific sound features under highly variable listening conditions and segregate distinct sound streams from multiple overlapping sources. The auditory system employs a variety of mechanisms to achieve this auditory scene analysis. First, neurons across levels of the auditory system exhibit compensatory adaptations to their gain and dynamic range in response to prevailing sound stimulus statistics in the environment. These adaptations allow for robust representations of sound features that are to a large degree invariant to the level of background noise. Second, listeners can selectively attend to a desired sound target in an environment with multiple sound sources. This selective auditory attention is another form of sensory gain control, enhancing the representation of an attended sound source while suppressing responses to unattended sounds. This review will examine both “bottom-up” gain alterations in response to changes in environmental sound statistics as well as “top-down” mechanisms that allow for selective extraction of specific sound features in a complex auditory scene. Finally, we will discuss how hearing loss interacts with these gain control mechanisms, and the adaptive and/or maladaptive perceptual consequences of this plasticity.

Paired Feed-Forward Excitation With Delayed Inhibition Allows High Frequency Computations Across Brain Regions

The transmission of high frequency temporal information across brain regions is critical to perception, but the mechanisms underlying such transmission remain unclear. Long-range projection patterns across brain areas are often comprised of paired feed-forward excitation followed closely by delayed inhibition, including the thalamic triad synapse, thalamic projections to cortex, and projections within the hippocampus. Previous studies have shown that these joint projections produce a shortened period of depolarization, sharpening the timing window over which the postsynaptic neuron can fire. Here we show that these projections can facilitate the transmission of high frequency computations even at frequencies that are highly filtered by neuronal membranes. This temporal facilitation occurred over a range of synaptic parameter values, including variations in synaptic strength, synaptic time constants, short-term synaptic depression, and the delay between excitation and inhibition. Further, these projections can coordinate computations across multiple network levels, even amid ongoing local activity. We suggest that paired feed-forward excitation and inhibition provide a hybrid signal-carrying both a value and a clock-like trigger-to allow circuits to be responsive to input whenever it arrives.

Feedforward mechanisms of cross-orientation interactions in mouse V1

Sensory neurons are modulated by context. For example, in mouse primary visual cortex (V1), neuronal responses to the preferred orientation are modulated by the presence of superimposed orientations ("plaids"). The effects of this modulation are diverse; some neurons are suppressed, while others have larger responses to a plaid than its components. We investigated whether this diversity could be explained by a unified circuit mechanism. We report that this masking is maintained during suppression of cortical activity, arguing against cortical mechanisms. Instead, the heterogeneity of plaid responses is explained by an interaction between stimulus geometry and orientation tuning. Highly selective neurons are uniformly suppressed by plaids, whereas the effects in weakly selective neurons depend on the spatial configuration of the stimulus, transitioning systematically between suppression and facilitation. Thus, the diverse responses emerge as a consequence of the spatial structure of feedforward inputs, with no need to invoke cortical interactions.

Sensory Adaptation in the Whisker-Mediated Tactile System: Physiology, Theory, and Function

In the natural environment, organisms are constantly exposed to a continuous stream of sensory input. The dynamics of sensory input changes with organism's behaviour and environmental context. The contextual variations may induce >100-fold change in the parameters of the stimulation that an animal experiences. Thus, it is vital for the organism to adapt to the new diet of stimulation. The response properties of neurons, in turn, dynamically adjust to the prevailing properties of sensory stimulation, a process known as "neuronal adaptation." Neuronal adaptation is a ubiquitous phenomenon across all sensory modalities and occurs at different stages of processing from periphery to cortex. In spite of the wealth of research on contextual modulation and neuronal adaptation in visual and auditory systems, the neuronal and computational basis of sensory adaptation in somatosensory system is less understood. Here, we summarise the recent finding and views about the neuronal adaptation in the rodent whisker-mediated tactile system and further summarise the functional effect of neuronal adaptation on the response dynamics and encoding efficiency of neurons at single cell and population levels along the whisker-mediated touch system in rodents. Based on direct and indirect pieces of evidence presented here, we suggest sensory adaptation provides context-dependent functional mechanisms for noise reduction in sensory processing, salience processing and deviant stimulus detection, shift between integration and coincidence detection, band-pass frequency filtering, adjusting neuronal receptive fields, enhancing neural coding and improving discriminability around adapting stimuli, energy conservation, and disambiguating encoding of principal features of tactile stimuli.

Steady-state measures of visual suppression

In the early visual system, suppression occurs between neurons representing different stimulus properties. This includes features such as orientation (cross-orientation suppression), eye-of-origin (interocular suppression) and spatial location (surround suppression), which are thought to involve distinct anatomical pathways. We asked if these separate routes to suppression can be differentiated by their pattern of gain control on the contrast response function measured in human participants using steady-state electroencephalography. Changes in contrast gain shift the contrast response function laterally, whereas changes in response gain scale the function vertically. We used a Bayesian hierarchical model to summarise the evidence for each type of gain control. A computational meta-analysis of 16 previous studies found the most evidence for contrast gain effects with overlaid masks, but no clear evidence favouring either response gain or contrast gain for other mask types. We then conducted two new experiments, comparing suppression from four mask types (monocular and dichoptic overlay masks, and aligned and orthogonal surround masks) on responses to sine wave grating patches flickering at 5Hz. At the occipital pole, there was strong evidence for contrast gain effects in all four mask types at the first harmonic frequency (5Hz). Suppression generally became stronger at more lateral electrode sites, but there was little evidence of response gain effects. At the second harmonic frequency (10Hz) suppression was stronger overall, and involved both contrast and response gain effects. Although suppression from different mask types involves distinct anatomical pathways, gain control processes appear to serve a common purpose, which we suggest might be to suppress less reliable inputs.

Adaptation supports short-term memory in a visual change detection task

The maintenance of short-term memories is critical for survival in a dynamically changing world. Previous studies suggest that this memory can be stored in the form of persistent neural activity or using a synaptic mechanism, such as with short-term plasticity. Here, we compare the predictions of these two mechanisms to neural and behavioral measurements in a visual change detection task. Mice were trained to respond to changes in a repeated sequence of natural images while neural activity was recorded using two-photon calcium imaging. We also trained two types of artificial neural networks on the same change detection task as the mice. Following fixed pre-processing using a pretrained convolutional neural network, either a recurrent neural network (RNN) or a feedforward neural network with short-term synaptic depression (STPNet) was trained to the same level of performance as the mice. While both networks are able to learn the task, the STPNet model contains units whose activity are more similar to the in vivo data and produces errors which are more similar to the mice. When images are omitted, an unexpected perturbation which was absent during training, mice often do not respond to the omission but are more likely to respond to the subsequent image. Unlike the RNN model, STPNet produces a similar pattern of behavior. These results suggest that simple neural adaptation mechanisms may serve as an important bottom-up memory signal in this task, which can be used by downstream areas in the decision-making process.

Animals have to adapt to environments with rich dynamics and maintain multiple types of memories. In this study, we focus on a visual change detection task in mice which requires short-term memory. Learning which features need to be maintained in short-term memory can be realized in a recurrent neural network by changing connections in the network, resulting in memory maintenance through persistent activity. However, in biological networks, a large diversity of time-dependent intrinsic mechanisms are also available. As an alternative to persistent neural activity, we find that learning to make use of internal adapting dynamics better matches both the observed neural activity and behavior of animals in this simple task. The presence of a large diversity of temporal traces could be one of the reasons for the diversity of cells observed. We believe that both learning to keep representations of relevant stimuli in persistent activity and learning to make use of intrinsic time-dependent mechanisms exist, and their relative use will be dependent on the exact task.

How auditory selectivity for sound timing arises: The diverse roles of GABAergic inhibition in shaping the excitation to interval-selective midbrain neurons

Across sensory systems, temporal frequency information is progressively transformed along ascending central pathways. Despite considerable effort to elucidate the mechanistic basis of these transformations, they remain poorly understood. Here we used a novel constellation of approaches, including whole-cell recordings and focal pharmacological manipulation, in vivo, and new computational algorithms that identify conductances resulting from excitation, inhibition and active membrane properties, to elucidate the mechanisms underlying the selectivity of midbrain auditory neurons for long temporal intervals. Surprisingly, we found that stimulus-driven excitation can be increased and its selectivity decreased following attenuation of inhibition with gabazine or intracellular delivery of fluoride. We propose that this nonlinear interaction is due to shunting inhibition. The rate-dependence of this inhibition results in the illusion that excitation to a cell shows greater temporal selectivity than is actually the case. We also show that rate-dependent depression of excitation, an important component of long-interval selectivity, can be decreased after attenuating inhibition. These novel findings indicate that nonlinear shunting inhibition plays a key role in shaping the amplitude and interval selectivity of excitation. Our findings provide a major advance in understanding how the brain decodes intervals and may explain paradoxical temporal selectivity of excitation to midbrain neurons reported previously.

A recurrent circuit implements normalization, simulating the dynamics of V1 activity

The normalization model has been applied to explain neural activity in diverse neural systems including primary visual cortex (V1). The model's defining characteristic is that the response of each neuron is divided by a factor that includes a weighted sum of activity of a pool of neurons. Despite the success of the normalization model, there are three unresolved issues. 1) Experimental evidence supports the hypothesis that normalization in V1 operates via recurrent amplification, i.e., amplifying weak inputs more than strong inputs. It is unknown how normalization arises from recurrent amplification. 2) Experiments have demonstrated that normalization is weighted such that each weight specifies how one neuron contributes to another's normalization pool. It is unknown how weighted normalization arises from a recurrent circuit. 3) Neural activity in V1 exhibits complex dynamics, including gamma oscillations, linked to normalization. It is unknown how these dynamics emerge from normalization. Here, a family of recurrent circuit models is reported, each of which comprises coupled neural integrators to implement normalization via recurrent amplification with arbitrary normalization weights, some of which can recapitulate key experimental observations of the dynamics of neural activity in V1.

Contrast-Dependence of Temporal Frequency Tuning in Mouse V1

The perception of speed is influenced by visual contrast. In primary visual cortex (V1), an early stage in the visual perception pathway, the neural tuning to speed is directly related to the neural tuning to temporal frequency of stimulus changes. The influence of contrast on speed perception can be caused by the joint dependency of neural responses in V1 on temporal frequency and contrast. Here, we investigated how tuning to contrast and temporal frequency in V1 of anesthetized mice are related. We found that temporal frequency tuning is contrast-dependent. V1 was more responsive at lower temporal frequencies than the dLGN, consistent with previous work at high contrast. The temporal frequency tuning moves toward higher temporal frequencies with increasing contrast. The low half-maximum temporal frequency does not change with contrast. The Heeger divisive normalization equation provides a good fit to many response characteristics in V1, but does not fit the dependency of temporal frequency and contrast with set of parameters for all temporal frequencies. Different mechanisms for normalization in the visual cortex may predict different relationships between temporal frequency and contrast non-linearity. Our data could help to make a model selection.

Contrast gain control occurs independently of both parvalbumin-positive interneuron activity and shunting inhibition in auditory cortex

Contrast gain control is the systematic adjustment of neuronal gain in response to the contrast of sensory input. It is widely observed in sensory cortical areas and has been proposed to be a canonical neuronal computation. Here, we investigated whether shunting inhibition from parvalbumin-positive interneurons-a mechanism involved in gain control in visual cortex-also underlies contrast gain control in auditory cortex. First, we performed extracellular recordings in the auditory cortex of anesthetized male mice and optogenetically manipulated the activity of parvalbumin-positive interneurons while varying the contrast of the sensory input. We found that both activation and suppression of parvalbumin interneuron activity altered the overall gain of cortical neurons. However, despite these changes in overall gain, we found that manipulating parvalbumin interneuron activity did not alter the strength of contrast gain control in auditory cortex. Furthermore, parvalbumin-positive interneurons did not show increases in activity in response to high-contrast stimulation, which would be expected if they drive contrast gain control. Finally, we performed in vivo whole-cell recordings in auditory cortical neurons during high- and low-contrast stimulation and found that no increase in membrane conductance was observed during high-contrast stimulation. Taken together, these findings indicate that while parvalbumin-positive interneuron activity modulates the overall gain of auditory cortical responses, other mechanisms are primarily responsible for contrast gain control in this cortical area.NEW & NOTEWORTHY We investigated whether contrast gain control is mediated by shunting inhibition from parvalbumin-positive interneurons in auditory cortex. We performed extracellular and intracellular recordings in mouse auditory cortex while presenting sensory stimuli with varying contrasts and manipulated parvalbumin-positive interneuron activity using optogenetics. We show that while parvalbumin-positive interneuron activity modulates the gain of cortical responses, this activity is not the primary mechanism for contrast gain control in auditory cortex.

Temporal information loss in the macaque early visual system

Stimuli that modulate neuronal activity are not always detectable, indicating a loss of information between the modulated neurons and perception. To identify where in the macaque visual system information about periodic light modulations is lost, signal-to-noise ratios were compared across simulated cone photoreceptors, lateral geniculate nucleus (LGN) neurons, and perceptual judgements. Stimuli were drifting, threshold-contrast Gabor patterns on a photopic background. The sensitivity of LGN neurons, extrapolated to populations, was similar to the monkeys' at low temporal frequencies. At high temporal frequencies, LGN sensitivity exceeded the monkeys' and approached the upper bound set by cone photocurrents. These results confirm a loss of high-frequency information downstream of the LGN. However, this loss accounted for only about 5% of the total. Phototransduction accounted for essentially all of the rest. Together, these results show that low temporal frequency information is lost primarily between the cones and the LGN, whereas high-frequency information is lost primarily within the cones, with a small additional loss downstream of the LGN.

Oscillatory recurrent gated neural integrator circuits (ORGaNICs), a unifying theoretical framework for neural dynamics

Working memory is an example of a cognitive and neural process that is not static but evolves dynamically with changing sensory inputs; another example is motor preparation and execution. We introduce a theoretical framework for neural dynamics, based on oscillatory recurrent gated neural integrator circuits (ORGaNICs), and apply it to simulate key phenomena of working memory and motor control. The model circuits simulate neural activity with complex dynamics, including sequential activity and traveling waves of activity, that manipulate (as well as maintain) information during working memory. The same circuits convert spatial patterns of premotor activity to temporal profiles of motor control activity and manipulate (e.g., time warp) the dynamics. Derivative-like recurrent connectivity, in particular, serves to manipulate and update internal models, an essential feature of working memory and motor execution. In addition, these circuits incorporate recurrent normalization, to ensure stability over time and robustness with respect to perturbations of synaptic weights.

Adaptation of the human auditory cortex to changing background noise

Speech communication in real-world environments requires adaptation to changing acoustic conditions. How the human auditory cortex adapts as a new noise source appears in or disappears from the acoustic scene remain unclear. Here, we directly measured neural activity in the auditory cortex of six human subjects as they listened to speech with abruptly changing background noises. We report rapid and selective suppression of acoustic features of noise in the neural responses. This suppression results in enhanced representation and perception of speech acoustic features. The degree of adaptation to different background noises varies across neural sites and is predictable from the tuning properties and speech specificity of the sites. Moreover, adaptation to background noise is unaffected by the attentional focus of the listener. The convergence of these neural and perceptual effects reveals the intrinsic dynamic mechanisms that enable a listener to filter out irrelevant sound sources in a changing acoustic scene.

Scaling Our World View: How Monoamines Can Put Context Into Brain Circuitry

Monoamines are presumed to be diffuse metabotropic neuromodulators of the topographically and temporally precise ionotropic circuitry which dominates CNS functions. Their malfunction is strongly implicated in motor and cognitive disorders, but their function in behavioral and cognitive processing is scarcely understood. In this paper, the principles of such a monoaminergic function are conceptualized for locomotor control. We find that the serotonergic system in the ventral spinal cord scales ionotropic signals and shows topographic order that agrees with differential gain modulation of ionotropic subcircuits. Whereas the subcircuits can collectively signal predictive models of the world based on life-long learning, their differential scaling continuously adjusts these models to changing mechanical contexts based on sensory input on a fast time scale of a few 100 ms. The control theory of biomimetic robots demonstrates that this precision scaling is an effective and resource-efficient solution to adapt the activation of individual muscle groups during locomotion to changing conditions such as ground compliance and carried load. Although it is not unconceivable that spinal ionotropic circuitry could achieve scaling by itself, neurophysiological findings emphasize that this is a unique functionality of metabotropic effects since recent recordings in sensorimotor circuitry conflict with mechanisms proposed for ionotropic scaling in other CNS areas. We substantiate that precision scaling of ionotropic subcircuits is a main functional principle for many monoaminergic projections throughout the CNS, implying that the monoaminergic circuitry forms a network within the network composed of the ionotropic circuitry. Thereby, we provide an early-level interpretation of the mechanisms of psychopharmacological drugs that interfere with the monoaminergic systems.

Mechanisms of Orientation Selectivity in the Primary Visual Cortex

The mechanisms underlying the emergence of orientation selectivity in the visual cortex have been, and continue to be, the subjects of intense scrutiny. Orientation selectivity reflects a dramatic change in the representation of the visual world: Whereas afferent thalamic neurons are generally orientation insensitive, neurons in the primary visual cortex (V1) are extremely sensitive to stimulus orientation. This profound change in the receptive field structure along the visual pathway has positioned V1 as a model system for studying the circuitry that underlies neural computations across the neocortex. The neocortex is characterized anatomically by the relative uniformity of its circuitry despite its role in processing distinct signals from region to region. A combination of physiological, anatomical, and theoretical studies has shed some light on the circuitry components necessary for generating orientation selectivity in V1. This targeted effort has led to critical insights, as well as controversies, concerning how neural circuits in the neocortex perform computations.

The divisive normalization model of V1 neurons: a comprehensive comparison of physiological data and model predictions

The physiological responses of simple and complex cells in the primary visual cortex (V1) have been studied extensively and modeled at different levels. At the functional level, the divisive normalization model (DNM; Heeger DJ. Vis Neurosci 9: 181-197, 1992) has accounted for a wide range of single-cell recordings in terms of a combination of linear filtering, nonlinear rectification, and divisive normalization. We propose standardizing the formulation of the DNM and implementing it in software that takes static grayscale images as inputs and produces firing rate responses as outputs. We also review a comprehensive suite of 30 empirical phenomena and report a series of simulation experiments that qualitatively replicate dozens of key experiments with a standard parameter set consistent with physiological measurements. This systematic approach identifies novel falsifiable predictions of the DNM. We show how the model simultaneously satisfies the conflicting desiderata of flexibility and falsifiability. Our key idea is that, while adjustable parameters are needed to accommodate the diversity across neurons, they must be fixed for a given individual neuron. This requirement introduces falsifiable constraints when this single neuron is probed with multiple stimuli. We also present mathematical analyses and simulation experiments that explicate some of these constraints.

Sensory noise predicts divisive reshaping of receptive fields

In order to respond reliably to specific features of their environment, sensory neurons need to integrate multiple incoming noisy signals. Crucially, they also need to compete for the interpretation of those signals with other neurons representing similar features. The form that this competition should take depends critically on the noise corrupting these signals. In this study we show that for the type of noise commonly observed in sensory systems, whose variance scales with the mean signal, sensory neurons should selectively divide their input signals by their predictions, suppressing ambiguous cues while amplifying others. Any change in the stimulus context alters which inputs are suppressed, leading to a deep dynamic reshaping of neural receptive fields going far beyond simple surround suppression. Paradoxically, these highly variable receptive fields go alongside and are in fact required for an invariant representation of external sensory features. In addition to offering a normative account of context-dependent changes in sensory responses, perceptual inference in the presence of signal-dependent noise accounts for ubiquitous features of sensory neurons such as divisive normalization, gain control and contrast dependent temporal dynamics.

Diverse effects of stimulus history in waking mouse auditory cortex

Responses to auditory stimuli are often strongly influenced by recent stimulus history. For example, in a paradigm called forward suppression, brief sounds can suppress the perception of, and the neural responses to, a subsequent sound, with the magnitude of this suppression depending on both the spectral and temporal distances between the sounds. As a step towards understanding the mechanisms that generate these adaptive representations in awake animals, we quantitatively characterize responses to two-tone sequences in the auditory cortex of waking mice. We find that cortical responses in a forward suppression paradigm are more diverse in waking mice than previously appreciated, that these responses vary between cells with different firing characteristics and waveform shapes, but that the variability in these responses is not substantially related to cortical depth or columnar location. Moreover, responses to the first tone in the sequence are often not linearly related to the suppression of the second tone response, suggesting that spike-frequency adaptation of cortical cells is not a large contributor to forward suppression or its variability. Instead, we use a simple multilayered model to show that cell-to-cell differences in the balance of intracortical inhibition and excitation will naturally produce such a diversity of forward interactions. We propose that diverse inhibitory connectivity allows the cortex to encode spectro-temporally fluctuating stimuli in multiple parallel ways.NEW & NOTEWORTHY Behavioral and neural responses to auditory stimuli are profoundly influenced by recent sounds, yet how this occurs is not known. Here, the authors show in the auditory cortex of awake mice that the quality of history-dependent effects is diverse and related to cell type, response latency, firing rates, and receptive field bandwidth. In a cortical model, differences in excitatory-inhibitory balance can produce this diversity, providing the cortex with multiple ways of representing temporally complex information.

Surround-Masking Affects Visual Estimation Ability

Visual estimation of numerosity involves the discrimination of magnitude between two distributions or perceptual sets that vary in number of elements. How performance on such estimation depends on peripheral sensory stimulation is unclear, even in typically developing adults. Here, we varied the central and surround contrast of stimuli that comprised a visual estimation task in order to determine whether mechanisms involved with the removal of unessential visual input functionally contributes toward number acuity. The visual estimation judgments of typically developed adults were significantly impaired for high but not low contrast surround stimulus conditions. The center and surround contrasts of the stimuli also differentially affected the accuracy of numerosity estimation depending on whether fewer or more dots were presented. Remarkably, observers demonstrated the highest mean percentage accuracy across stimulus conditions in the discrimination of more elements when the surround contrast was low and the background luminance of the central region containing the elements was dark (black center). Conversely, accuracy was severely impaired during the discrimination of fewer elements when the surround contrast was high and the background luminance of the central region was mid level (gray center). These findings suggest that estimation ability is functionally related to the quality of low-order filtration of unessential visual information. These surround masking results may help understanding of the poor visual estimation ability commonly observed in developmental dyscalculia.

Functional implications of orientation maps in primary visual cortex

Stimulus orientation in the primary visual cortex of primates and carnivores is mapped as iso-orientation domains radiating from pinwheel centres, where orientation preferences of neighbouring cells change circularly. Whether this orientation map has a function is currently debated, because many mammals, such as rodents, do not have such maps. Here we show that two fundamental properties of visual cortical responses, contrast saturation and cross-orientation suppression, are stronger within cat iso-orientation domains than at pinwheel centres. These differences develop when excitation (not normalization) from neighbouring oriented neurons is applied to different cortical orientation domains and then balanced by inhibition from un-oriented neurons. The functions of the pinwheel mosaic emerge from these local intra-cortical computations: Narrower tuning, greater cross-orientation suppression and higher contrast gain of iso-orientation cells facilitate extraction of object contours from images, whereas broader tuning, greater linearity and less suppression of pinwheel cells generate selectivity for surface patterns and textures.

Stimulus orientation in the primary visual cortex of primates and carnivores is mapped into a geometrical mosaic but the functional implications of these maps remain debated. Here the authors reveal an association between the structure of cortical orientation maps in cats, and the functions of local cortical circuits in processing patterns and contours.

Network Receptive Field Modeling Reveals Extensive Integration and Multi-feature Selectivity in Auditory Cortical Neurons

Cortical sensory neurons are commonly characterized using the receptive field, the linear dependence of their response on the stimulus. In primary auditory cortex neurons can be characterized by their spectrotemporal receptive fields, the spectral and temporal features of a sound that linearly drive a neuron. However, receptive fields do not capture the fact that the response of a cortical neuron results from the complex nonlinear network in which it is embedded. By fitting a nonlinear feedforward network model (a network receptive field) to cortical responses to natural sounds, we reveal that primary auditory cortical neurons are sensitive over a substantially larger spectrotemporal domain than is seen in their standard spectrotemporal receptive fields. Furthermore, the network receptive field, a parsimonious network consisting of 1-7 sub-receptive fields that interact nonlinearly, consistently better predicts neural responses to auditory stimuli than the standard receptive fields. The network receptive field reveals separate excitatory and inhibitory sub-fields with different nonlinear properties, and interaction of the sub-fields gives rise to important operations such as gain control and conjunctive feature detection. The conjunctive effects, where neurons respond only if several specific features are present together, enable increased selectivity for particular complex spectrotemporal structures, and may constitute an important stage in sound recognition. In conclusion, we demonstrate that fitting auditory cortical neural responses with feedforward network models expands on simple linear receptive field models in a manner that yields substantially improved predictive power and reveals key nonlinear aspects of cortical processing, while remaining easy to interpret in a physiological context.

Spatiotemporal properties of microsaccades: Model predictions and experimental tests

Microsaccades are involuntary and very small eye movements during fixation. Recently, the microsaccade-related neural dynamics have been extensively investigated both in experiments and by constructing neural network models. Experimentally, microsaccades also exhibit many behavioral properties. It's well known that the behavior properties imply the underlying neural dynamical mechanisms, and so are determined by neural dynamics. The behavioral properties resulted from neural responses to microsaccades, however, are not yet understood and are rarely studied theoretically. Linking neural dynamics to behavior is one of the central goals of neuroscience. In this paper, we provide behavior predictions on spatiotemporal properties of microsaccades according to microsaccade-induced neural dynamics in a cascading network model, which includes both retinal adaptation and short-term depression (STD) at thalamocortical synapses. We also successfully give experimental tests in the statistical sense. Our results provide the first behavior description of microsaccades based on neural dynamics induced by behaving activity, and so firstly link neural dynamics to behavior of microsaccades. These results indicate strongly that the cascading adaptations play an important role in the study of microsaccades. Our work may be useful for further investigations of the microsaccadic behavioral properties and of the underlying neural dynamical mechanisms responsible for the behavioral properties.

The neurogenesis of P1 and N1: A concurrent EEG/LFP study

It is generally recognised that event related potentials (ERPs) of electroencephalogram (EEG) primarily reflect summed post-synaptic activity of the local pyramidal neural population(s). However, it is still not understood how the positive and negative deflections (e.g. P1, N1 etc) observed in ERP recordings are related to the underlying excitatory and inhibitory post-synaptic activity. We investigated the neurogenesis of P1 and N1 in ERPs by pharmacologically manipulating inhibitory post-synaptic activity in the somatosensory cortex of rodent, and concurrently recording EEG and local field potentials (LFPs). We found that the P1 wave in the ERP and LFP of the supragranular layers is determined solely by the excitatory post-synaptic activity of the local pyramidal neural population, as is the initial segment of the N1 wave across cortical depth. The later part of the N1 wave was modulated by inhibitory post-synaptic activity, with its peak and the pulse width increasing as inhibition was reduced. These findings suggest that the temporal delay of inhibition with respect to excitation observed in intracellular recordings is also reflected in extracellular field potentials (FPs), resulting in a temporal window during which only excitatory post-synaptic activity and leak channel activity are recorded in the ERP and evoked LFP time series. Based on these findings, we provide clarification on the interpretation of P1 and N1 in terms of the excitatory and inhibitory post-synaptic activities of the local pyramidal neural population(s).

A Computational Framework for Realistic Retina Modeling

Computational simulations of the retina have led to valuable insights about the biophysics of its neuronal activity and processing principles. A great number of retina models have been proposed to reproduce the behavioral diversity of the different visual processing pathways. While many of these models share common computational stages, previous efforts have been more focused on fitting specific retina functions rather than generalizing them beyond a particular model. Here, we define a set of computational retinal microcircuits that can be used as basic building blocks for the modeling of different retina mechanisms. To validate the hypothesis that similar processing structures may be repeatedly found in different retina functions, we implemented a series of retina models simply by combining these computational retinal microcircuits. Accuracy of the retina models for capturing neural behavior was assessed by fitting published electrophysiological recordings that characterize some of the best-known phenomena observed in the retina: adaptation to the mean light intensity and temporal contrast, and differential motion sensitivity. The retinal microcircuits are part of a new software platform for efficient computational retina modeling from single-cell to large-scale levels. It includes an interface with spiking neural networks that allows simulation of the spiking response of ganglion cells and integration with models of higher visual areas.

Exploring the Use of Sensorial LTP/LTD-Like Stimulation to Modulate Human Performance for Complex Visual Stimuli

Is it possible to passively induce visual learning/unlearning in humans for complex stimuli such as faces? We addressed this question in a series of behavioral studies using passive visual stimulation (flickering of faces at specific temporal frequencies) inspired by well-known synaptic mechanisms of learning: long-term potentiation (LTP) vs long-term depression (LTD). We administered a face identity change detection task before and after a passive stimulation protocol to test for potential changes in visual performance. First, with bilateral stimulation, subjects undergoing high-frequency LTP-like stimulation outperformed those submitted to low-frequency LTD-like stimulation despite equivalent baseline performance (exp. 1). Second, unilateral stimulation replicated the differential modulation of performance, but in a hemifield-specific way (exp. 2). Third, for both stimulation groups, a sudden temporary drop in performance on the stimulated side immediately after the stimulation, followed by progressive recovering, can suggest either 'visual fatigue' or 'face adaptation' effects due to the stimulation. Fourth, we tested the life-time of these modulatory effects, revealing they vanish after one hour delay (exp. 3). Fifth, a control study (exp. 4) using low-level visual stimuli also failed to show longer-term effects of sensory stimulation, despite reports of strong effects in the literature. Future studies should determine the necessary and sufficient conditions enabling robust long-term modulation of visual performance using this technique. This step is required to consider further use in fundamental research (e.g., to study neural circuits involved in selective visual processing) and potential educational or clinical applications (e.g., inhibiting socially-irrelevant aspects of face processing in autism).

Polarity-Specific Cortical Effects of Transcranial Direct Current Stimulation in Primary Somatosensory Cortex of Healthy Humans

Transcranial direct current stimulation (tDCS) is a non-invasive stimulation method that has been shown to modulate the excitability of the motor and visual cortices in human subjects in a polarity dependent manner in previous studies. The aim of our study was to investigate whether anodal and cathodal tDCS can also be used to modulate the excitability of the human primary somatosensory cortex (S1). We measured paired-pulse suppression (PPS) of somatosensory evoked potentials in 36 right-handed volunteers before and after anodal, cathodal, or sham stimulation over the right non-dominant S1. Paired-pulse stimulation of the median nerve was performed at the dominant and non-dominant hand. After anodal tDCS, PPS was reduced in the ipsilateral S1 compared to sham stimulation, indicating an excitatory effect of anodal tDCS. In contrast, PPS in the stimulated left hemisphere was increased after cathodal tDCS, indicating an inhibitory effect of cathodal tDCS. Sham stimulation induced no pre-post differences. Thus, tDCS can be used to modulate the excitability of S1 in polarity-dependent manner, which can be assessed by PPS. An interesting topic for further studies could be the investigation of direct correlations between sensory changes and excitability changes induced by tDCS.

A single dose of lorazepam reduces paired-pulse suppression of median nerve evoked somatosensory evoked potentials

Paired-pulse behaviour in the somatosensory cortex is an approach to obtain insights into cortical processing modes and to obtain markers of changes of cortical excitability attributable to learning or pathological states. Numerous studies have demonstrated suppression of the response to the stimulus that follows a first one after a short interval, but the underlying mechanisms remain elusive, although there is agreement that GABAergic mechanisms seem to play a crucial role. We therefore aimed to explore the influence of the GABAA agonist lorazepam on paired-pulse somatosensory evoked potentials (SEPs). We recorded and analysed SEPs after paired median nerve stimulation in healthy individuals before and after they had received a single dose of 2.5 mg of lorazepam as compared with a control group receiving placebo. Paired-pulse suppression was expressed as a ratio of the amplitudes of the second and the first peaks. We found that, after lorazepam application, paired-pulse suppression of the cortical N20 component remained unchanged, but suppression of the N20-P25 complex was significantly reduced, indicative of GABAergic involvement in intracortical processing. Our data suggest that lorazepam most likely enhances inhibition within the cortical network of interneurons responsible for creating paired-pulse suppression, leading to reduced inhibitory drive with a subsequently reduced amount of suppression. The results provide further evidence that GABAA -mediated mechanisms are involved in the generation of median nerve evoked paired-pulse suppression.

Fast response and high sensitivity to microsaccades in a cascading-adaptation neural network with short-term synaptic depression

Microsaccades are very small eye movements during fixation. Experimentally, they have been found to play an important role in visual information processing. However, neural responses induced by microsaccades are not yet well understood and are rarely studied theoretically. Here we propose a network model with a cascading adaptation including both retinal adaptation and short-term depression (STD) at thalamocortical synapses. In the neural network model, we compare the microsaccade-induced neural responses in the presence of STD and those without STD. It is found that the cascading with STD can give rise to faster and sharper responses to microsaccades. Moreover, STD can enhance response effectiveness and sensitivity to microsaccadic spatiotemporal changes, suggesting improved detection of small eye movements (or moving visual objects). We also explore the mechanism of the response properties in the model. Our studies strongly indicate that STD plays an important role in neural responses to microsaccades. Our model considers simultaneously retinal adaptation and STD at thalamocortical synapses in the study of microsaccade-induced neural activity, and may be useful for further investigation of the functional roles of microsaccades in visual information processing.

SYNAPTIC DEPRESSION IN DEEP NEURAL NETWORKS FOR SPEECH PROCESSING

A characteristic property of biological neurons is their ability to dynamically change the synaptic efficacy in response to variable input conditions. This mechanism, known as synaptic depression, significantly contributes to the formation of normalized representation of speech features. Synaptic depression also contributes to the robust performance of biological systems. In this paper, we describe how synaptic depression can be modeled and incorporated into deep neural network architectures to improve their generalization ability. We observed that when synaptic depression is added to the hidden layers of a neural network, it reduces the effect of changing background activity in the node activations. In addition, we show that when synaptic depression is included in a deep neural network trained for phoneme classification, the performance of the network improves under noisy conditions not included in the training phase. Our results suggest that more complete neuron models may further reduce the gap between the biological performance and artificial computing, resulting in networks that better generalize to novel signal conditions.

Chronic recordings reveal tactile stimuli can suppress spontaneous activity of neurons in somatosensory cortex of awake and anesthetized primates

In somatosensory cortex, tactile stimulation within the neuronal receptive field (RF) typically evokes a transient excitatory response with or without postexcitatory inhibition. Here, we describe neuronal responses in which stimulation on the hand is followed by suppression of the ongoing discharge. With the use of 16-channel microelectrode arrays implanted in the hand representation of primary somatosensory cortex of New World monkeys and prosimian galagos, we recorded neuronal responses from single units and neuron clusters. In 66% of our sample, neuron activity tended to display suppression of firing when regions of skin outside of the excitatory RF were stimulated. In a small proportion of neurons, single-site indentations suppressed firing without initial increases in response to any of the tested sites on the hand. Latencies of suppressive responses to skin indentation (usually 12-34 ms) were similar to excitatory response latencies. The duration of inhibition varied across neurons. Although most observations were from anesthetized animals, we also found similar neuron response properties in one awake galago. Notably, suppression of ongoing neuronal activity did not require conditioning stimuli or multi-site stimulation. The suppressive effects were generally seen following single-site skin indentations outside of the neuron's minimal RF and typically on different digits and palm pads, which have not often been studied in this context. Overall, the characteristics of widespread suppressive or inhibitory response properties with and without initial facilitative or excitatory responses add to the growing evidence that neurons in primary somatosensory cortex provide essential processing for integrating sensory stimulation from across the hand.

Mathematical impairment associated with high-contrast abnormalities in change detection and magnocellular visual evoked response

The cause of developmental dyscalculia, a specific deficit in acquisition of arithmetic skills, particularly of enumeration, has never been investigated with respect to the patency of the visual magnocellular system. Here, the question of dysfunction of the afferent magnocellular cortical input and its dorsal stream projections was tested directly using nonlinear analysis of the visual evoked potential (VEP) and through the psychophysical ability to rapidly detect visual change. A group of young adults with self-reported deficiencies of arithmetical ability, showed marked impairment in magnitude estimation and enumeration performance-though not in lexical decision reaction times when compared with an arithmetically capable group controlled for age and handedness. Multifocal nonlinear VEPs were recorded at low (24 %) and high (96 %) contrast. First- and second-order VEP kernels were comparable between groups at low contrast, but not at high contrast. The mathematically impaired group showed an abnormal lack of contrast saturation in the shortest latency first-order peak (N60) and a delayed P100 positivity in the first slice of the second-order kernel. Both features have previously been argued to be physiological markers of magnocellular function. Mathematically impaired participants also performed worse on a gap paradigm change detection for digit task showing increased reaction times for high-contrast stimuli but not for low-contrast stimuli compared with controls. The VEP results give direct evidence of abnormality in the occipital processing of magnocellular information in those with mathematical impairment. The anomalous high visual contrast physiological and psychophysical performance suggests an abnormality in the inhibitory processes that normally result in saturation of contrast gain in the magnocellular system.

Probing the involvement of the earliest levels of cortical processing in motion extrapolation with rapid forms of visual motion priming and adaptation

In this study, we investigated the effect of brief motion priming and adaptation, occurring at the earliest levels of the cortical visual stream, on time-to-contact (TTC) estimation of a target passing behind an occluder. By using different exposure times of directional motion presented in the occluder area prior to the target's disappearance behind it, our aim was to modulate (prime or adapt) the extrapolated motion of the invisible target, thus producing different TTC estimates. Our results showed that longer (yet subsecond) exposures to motion in the same direction as the target produced late TTC estimates, whereas shorter exposures produced shorter TTC estimates, indicating that rapid forms of motion adaptation and motion priming affect extrapolated motion. Our findings suggest that motion extrapolation might occur at the earliest levels of cortical processing of motion, at which these rapid mechanisms of priming and adaptation take place.

Acute alcohol exposure impairs neural representation of visual motion speed in the visual cortex area posteromedial lateral suprasylvian cortex of cats

BACKGROUND

Psychophysical and behavioral studies have demonstrated that perception of motion can be impaired by acute alcohol exposure. The neural activities of posteromedial lateral suprasylvian cortex (PMLS) of cats are directly linked to the perception of visual motion speed. To date, there have been no studies on the effects of acute alcohol exposure in vivo upon the representation of speed in PMLS neurons.

METHODS

Alcohol was administered intravenously as a 20% (v/v) saline solution via a syringe at a dose levels of 0.5, 1, or 2 g/kg to generate a series of blood alcohol concentrations. Using extracellular single-unit recording technique, we recorded the speed-tuning properties of PMLS neurons that responded to random-dot patterns before and after alcohol administration, and simultaneously monitored the concentration of ethanol by detecting the breath alcohol concentration using a breath analyzer.

RESULTS

After acute alcohol treatment, PMLS cells preferred lower speeds. A broadened speed-tuning bandwidth of PMLS cells was also observed after acute alcohol administration. Additionally, response modulation and discriminative capacity for speed of visual motion in the PMLS cells were significantly impaired after acute alcohol exposure. Concurrently, PMLS cells after acute alcohol exposure showed decreased spontaneous activity, peak responses, and signal-to-noise ratios.

CONCLUSIONS

There is a significant functional degradation in the neural representation of visual motion speed in PMLS of cats after acute alcohol exposure. These neural changes may contribute to the alcohol-related deficits in visual motion perception observed in behavioral studies.

Contextual modulation and stimulus selectivity in extrastriate cortex

Contextual modulation is observed throughout the visual system, using techniques ranging from single-neuron recordings to behavioral experiments. Its role in generating feature selectivity within the retina and primary visual cortex has been extensively described in the literature. Here, we describe how similar computations can also elaborate feature selectivity in the extrastriate areas of both the dorsal and ventral streams of the primate visual system. We discuss recent work that makes use of normalization models to test specific roles for contextual modulation in visual cortex function. We suggest that contextual modulation renders neuronal populations more selective for naturalistic stimuli. Specifically, we discuss contextual modulation's role in processing optic flow in areas MT and MST and for representing naturally occurring curvature and contours in areas V4 and IT. We also describe how the circuitry that supports contextual modulation is robust to variations in overall input levels. Finally, we describe how this theory relates to other hypothesized roles for contextual modulation.

The effects of short-term synaptic depression at thalamocortical synapses on orientation tuning in cat V1

We examine the effects of short-term synaptic depression on the orientation tuning of the LGN input to simple cells in cat primary visual cortex (V1). The total LGN input has an untuned component as well as a tuned component, both of which grow with stimulus contrast. The untuned component is not visible in the firing rate responses of the simple cells. The suppression of the contribution of the untuned input component to firing rate responses is key to establishing orientation selectivity and its invariance with stimulus contrast. It has been argued that synaptic depression of LGN inputs could contribute to the selective suppression of the untuned component and thus contribute to the tuning observed in simple cells. We examine this using a model fit to the depression observed at thalamocortical synapses in-vivo, and compare this to an earlier model fit based on in-vitro observations. We examine the tuning of both the conductance and the firing rate induced in simple cells by the net LGN input. We find that depression causes minimal suppression of the untuned component. The primary effect of depression is to cause the contrast response curve to saturate at lower contrasts without differentially affecting the tuned vs. untuned components. This effect is slightly weaker for in-vivo vs. in-vitro parameters. Thus, synaptic depression of LGN inputs does not appreciably contribute to the orientation tuning of V1 simple cells.

Visual nonclassical receptive field effects emerge from sparse coding in a dynamical system

Extensive electrophysiology studies have shown that many V1 simple cells have nonlinear response properties to stimuli within their classical receptive field (CRF) and receive contextual influence from stimuli outside the CRF modulating the cell's response. Models seeking to explain these non-classical receptive field (nCRF) effects in terms of circuit mechanisms, input-output descriptions, or individual visual tasks provide limited insight into the functional significance of these response properties, because they do not connect the full range of nCRF effects to optimal sensory coding strategies. The (population) sparse coding hypothesis conjectures an optimal sensory coding approach where a neural population uses as few active units as possible to represent a stimulus. We demonstrate that a wide variety of nCRF effects are emergent properties of a single sparse coding model implemented in a neurally plausible network structure (requiring no parameter tuning to produce different effects). Specifically, we replicate a wide variety of nCRF electrophysiology experiments (e.g., end-stopping, surround suppression, contrast invariance of orientation tuning, cross-orientation suppression, etc.) on a dynamical system implementing sparse coding, showing that this model produces individual units that reproduce the canonical nCRF effects. Furthermore, when the population diversity of an nCRF effect has also been reported in the literature, we show that this model produces many of the same population characteristics. These results show that the sparse coding hypothesis, when coupled with a biophysically plausible implementation, can provide a unified high-level functional interpretation to many response properties that have generally been viewed through distinct mechanistic or phenomenological models.

Simple cells in the primary visual cortex (V1) demonstrate many response properties that are either nonlinear or involve response modulations (i.e., stimuli that do not cause a response in isolation alter the cell's response to other stimuli). These non-classical receptive field (nCRF) effects are generally modeled individually and their collective role in biological vision is not well understood. Previous work has shown that classical receptive field (CRF) properties of V1 cells (i.e., the spatial structure of the visual field responsive to stimuli) could be explained by the sparse coding hypothesis, which is an optimal coding model that conjectures a neural population should use the fewest number of cells simultaneously to represent each stimulus. In this paper, we have performed extensive simulated physiology experiments to show that many nCRF response properties are simply emergent effects of a dynamical system implementing this same sparse coding model. These results suggest that rather than representing disparate information processing operations themselves, these nCRF effects could be consequences of an optimal sensory coding strategy that attempts to represent each stimulus most efficiently. This interpretation provides a potentially unifying high-level functional interpretation to many response properties that have generally been viewed through distinct models.

Mechanisms of neuronal computation in mammalian visual cortex

Orientation selectivity in the primary visual cortex (V1) is a receptive field property that is at once simple enough to make it amenable to experimental and theoretical approaches and yet complex enough to represent a significant transformation in the representation of the visual image. As a result, V1 has become an area of choice for studying cortical computation and its underlying mechanisms. Here we consider the receptive field properties of the simple cells in cat V1--the cells that receive direct input from thalamic relay cells--and explore how these properties, many of which are highly nonlinear, arise. We have found that many receptive field properties of V1 simple cells fall directly out of Hubel and Wiesel's feedforward model when the model incorporates realistic neuronal and synaptic mechanisms, including threshold, synaptic depression, response variability, and the membrane time constant.