Inferring nonlinear neuronal computation based on physiologically plausible inputs

A systematic analysis of the joint effects of ganglion cells, lagged LGN cells, and intercortical inhibition on spatiotemporal processing and direction selectivity

Simple cells in the visual cortex process spatial as well as temporal information of the visual stream and enable the perception of motion information. Previous work suggests different mechanisms associated with direction selectivity, such as a temporal offset in thalamocortical input stream through lagged and non-lagged cells of the lateral geniculate nucleus (LGN), or solely from intercortical inhibition, or through a baseline selectivity provided by the thalamocortical connection tuned by intercortical inhibition. While there exists a large corpus of models for spatiotemporal receptive fields, the majority of them built-in the spatiotemporal dynamics by utilizing a combination of spatial and temporal functions and thus, do not explain the emergence of spatiotemporal dynamics on basis of network dynamics emerging in the retina and the LGN. In order to better comprehend the emergence of spatiotemporal processing and direction selectivity, we used a spiking neural network to implement the visual pathway from the retina to the primary visual cortex. By varying different functional parts in our network, we demonstrate how the direction selectivity of simple cells emerges through the interplay between two components: tuned intercortical inhibition and a temporal offset in the feedforward path through lagged LGN cells. In contrast to previous findings, our model simulations suggest an alternative dynamic between these two mechanisms: While intercortical inhibition alone leads to bidirectional selectivity, a temporal shift in the thalamocortical pathway breaks this symmetry in favor of one direction, leading to unidirectional selectivity.

Filter-based models of suppression in retinal ganglion cells: Comparison and generalization across species and stimuli

The dichotomy of excitation and suppression is one of the canonical mechanisms explaining the complexity of neural activity. Computational models of the interplay of excitation and suppression in single neurons aim at investigating how this interaction affects a neuron's spiking responses and shapes the encoding of sensory stimuli. Here, we compare the performance of three filter-based stimulus-encoding models for predicting retinal ganglion cell responses recorded from axolotl, mouse, and marmoset retina to different types of temporally varying visual stimuli. Suppression in these models is implemented via subtractive or divisive interactions of stimulus filters or by a response-driven feedback module. For the majority of ganglion cells, the subtractive and divisive models perform similarly and outperform the feedback model as well as a linear-nonlinear (LN) model with no suppression. Comparison between the subtractive and the divisive model depends on cell type, species, and stimulus components, with the divisive model generalizing best across temporal stimulus frequencies and visual contrast and the subtractive model capturing in particular responses for slow temporal stimulus dynamics and for slow axolotl cells. Overall, we conclude that the divisive and subtractive models are well suited for capturing interactions of excitation and suppression in ganglion cells and perform best for different temporal regimes of these interactions.

Characterization of receptive fields in the dorsal lateral geniculate nucleus of the tammar wallaby

Orientation selectivity is a prominent feature of neurons in the mammalian primary visual cortex (V1), yet its emergence along the visual pathway varies across species. In carnivores and primates, neurons with elongated and orientation-selective receptive fields (RFs) emerge in V1, whereas in mice such RFs appear earlier, in the retina or dorsal lateral geniculate nucleus (dLGN). Here, we investigate the RF properties of neurons in the dLGN of a marsupial, the wallaby (Macropus eugenii) (n = 2; males), using multichannel electrodes and nonlinear input model (NIM) analysis. Do dLGN RFs resemble those of carnivores and primates or exhibit unique characteristics, particularly regarding orientation selectivity? We found that 82% of neurons have a predominant ON-center response. We identified two main cell types: X-cells (n = 15/22), which exhibit linear properties, and Y-cells (n = 7/22), which display nonlinear characteristics. Most dLGN RFs were blob-like and lacked the oriented structures seen in cortical neurons but some had slightly elongated central areas. These results indicate that robust orientation selectivity develops fully in V1 (76% of neurons). However, mild orientation biases were observed in 41% of dLGN neurons. This study enhances our understanding of visual processing in marsupials and underscores the evolutionary significance of orientation selectivity in mammalian visual pathways.NEW & NOTEWORTHY This study examines receptive field (RF) properties of neurons in the dorsal lateral geniculate nucleus (dLGN) of wallabies using multichannel electrodes and nonlinear input model (NIM) analysis. We identified two main cell types: X-cells (linear) and Y-cells (nonlinear). Most dLGN RFs were blob-like, with mild orientation biases in 41% of neurons, indicating robust orientation selectivity develops fully in primary visual cortex (V1) (76% of neurons).

Optimal Estimation of Local Motion-in-Depth with Naturalistic Stimuli

Estimating the motion of objects in depth is important for behavior and is strongly supported by binocular visual cues. To understand both how the brain should estimate motion in depth and how natural constraints shape and limit performance in two local 3D motion tasks, we develop image-computable ideal observers from a large number of binocular video clips created from a dataset of natural images. The observers spatiotemporally filter the videos and nonlinearly decode 3D motion from the filter responses. The optimal filters and decoder are dictated by the task-relevant image statistics and are specific to each task. Multiple findings emerge. First, two distinct filter subpopulations are spontaneously learned for each task. For 3D speed estimation, filters emerge for processing either changing disparities over time or interocular velocity differences, cues that are used by humans. For 3D direction estimation, filters emerge for discriminating either left-right or toward-away motion. Second, the filter responses, conditioned on the latent variable, are well-described as jointly Gaussian, and the covariance of the filter responses carries the information about the task-relevant latent variable. Quadratic combination is thus necessary for optimal decoding, which can be implemented by biologically plausible neural computations. Finally, the ideal observer yields nonobvious-and in some cases counterintuitive-patterns of performance like those exhibited by humans. Important characteristics of human 3D motion processing and estimation may therefore result from optimal information processing in the early visual system.

Feature selectivity and invariance in marsupial primary visual cortex

A fundamental question in sensory neuroscience revolves around how neurons represent complex visual stimuli. In mammalian primary visual cortex (V1), neurons decode intricate visual features to identify objects, with most being selective for edge orientation, but with half of those also developing invariance to edge position within their receptive fields. Position invariance allows cells to continue to code an edge even when it moves around. Combining feature selectivity and invariance is integral to successful object recognition. Considering the marsupial-eutherian divergence 160 million years ago, we explored whether feature selectivity and invariance was similar in marsupials and eutherians. We recovered the spatial filters and non-linear processing characteristics of the receptive fields of neurons in wallaby V1 and compared them with previous results from cat cortex. We stimulated the neurons in V1 with white Gaussian noise and analysed responses using the non-linear input model. Wallabies exhibit the same high percentage of orientation selective neurons as cats. However, in wallabies we observed a notably higher prevalence of neurons with three or more filters compared to cats. We show that having three or more filters substantially increases phase invariance in the V1s of both species, but that wallaby V1 accentuates this feature, suggesting that the species condenses more processing into the earliest cortical stage. These findings suggest that evolution has led to more than one solution to the problem of creating complex visual processing strategies. KEY POINTS: Previous studies have shown that the primary visual cortex (V1) in mammals is essential for processing complex visual stimuli, with neurons displaying selectivity for edge orientation and position. This research explores whether the visual processing mechanisms in marsupials, such as wallabies, are similar to those in eutherian mammals (e.g. cats). The study found that wallabies have a higher prevalence of neurons with multiple spatial filters in V1, indicating more complex visual processing. Using a non-linear input model, we demonstrated that neurons with three or more filters increase phase invariance. These findings suggest that marsupials and eutherian mammals have evolved similar strategies for visual processing, but marsupials have condensed more capacity to build phase invariance into the first step in the cortical pathway.

Feature-specific divisive normalization improves natural image encoding for depth perception

Vision science and visual neuroscience seek to understand how stimulus and sensor properties limit the precision with which behaviorally-relevant latent variables are encoded and decoded. In the primate visual system, binocular disparity-the canonical cue for stereo-depth perception-is initially encoded by a set of binocular receptive fields with a range of spatial frequency preferences. Here, with a stereo-image database having ground-truth disparity information at each pixel, we examine how response normalization and receptive field properties determine the fidelity with which binocular disparity is encoded in natural scenes. We quantify encoding fidelity by computing the Fisher information carried by the normalized receptive field responses. Several findings emerge from an analysis of the response statistics. First, broadband (or feature-unspecific) normalization yields Laplace-distributed receptive field responses, and narrowband (or feature-specific) normalization yields Gaussian-distributed receptive field responses. Second, the Fisher information in narrowband-normalized responses is larger than in broadband-normalized responses by a scale factor that grows with population size. Third, the most useful spatial frequency decreases with stimulus size and the range of spatial frequencies that is useful for encoding a given disparity decreases with disparity magnitude, consistent with neurophysiological findings. Fourth, the predicted patterns of psychophysical performance, and absolute detection threshold, match human performance with natural and artificial stimuli. The current computational efforts establish a new functional role for response normalization, and bring us closer to understanding the principles that should govern the design of neural systems that support perception in natural scenes.

Applying Super-Resolution and Tomography Concepts to Identify Receptive Field Subunits in the Retina

Spatially nonlinear stimulus integration by retinal ganglion cells lies at the heart of various computations performed by the retina. It arises from the nonlinear transmission of signals that ganglion cells receive from bipolar cells, which thereby constitute functional subunits within a ganglion cell's receptive field. Inferring these subunits from recorded ganglion cell activity promises a new avenue for studying the functional architecture of the retina. This calls for efficient methods, which leave sufficient experimental time to leverage the acquired knowledge for further investigating identified subunits. Here, we combine concepts from super-resolution microscopy and computed tomography and introduce super-resolved tomographic reconstruction (STR) as a technique to efficiently stimulate and locate receptive field subunits. Simulations demonstrate that this approach can reliably identify subunits across a wide range of model variations, and application in recordings of primate parasol ganglion cells validates the experimental feasibility. STR can potentially reveal comprehensive subunit layouts within only a few tens of minutes of recording time, making it ideal for online analysis and closed-loop investigations of receptive field substructure in retina recordings.

Data-driven modelling of visual receptive fields: comparison between the generalized quadratic model and the nonlinear input model

Objective: Neurons in primary visual cortex (V1) display a range of sensitivity in their response to translations of their preferred visual features within their receptive field: from high specificity to a precise position through to complete invariance. This visual feature selectivity and invariance is frequently modeled by applying a selection of linear spatial filters to the input image, that define the feature selectivity, followed by a nonlinear function that combines the filter outputs, that defines the invariance, to predict the neural response. We compare two such classes of model, that are both popular and parsimonious, the generalized quadratic model (GQM) and the nonlinear input model (NIM). These two classes of model differ primarily in that the NIM can accommodate a greater diversity in the form of nonlinearity that is applied to the outputs of the filters.Approach: We compare the two model types by applying them to data from multielectrode recordings from cat primary visual cortex in response to spatially white Gaussian noise After fitting both classes of model to a database of 342 single units (SUs), we analyze the qualitative and quantitative differences in the visual feature processing performed by the two models and their ability to predict neural response.Main results: We find that the NIM predicts response rates on a held-out data at least as well as the GQM for 95% of SUs. Superior performance occurs predominantly for those units with above average spike rates and is largely due to the NIMs ability to capture aspects of the model's nonlinear function cannot be captured with the GQM rather than differences in the visual features being processed by the two different models.Significance: These results can help guide model choice for data-driven receptive field modelling.

Estimating receptive fields of simple and complex cells in early visual cortex: A convolutional neural network model with parameterized rectification

Neurons in the primary visual cortex respond selectively to simple features of visual stimuli, such as orientation and spatial frequency. Simple cells, which have phase-sensitive responses, can be modeled by a single receptive field filter in a linear-nonlinear model. However, it is challenging to analyze phase-invariant complex cells, which require more elaborate models having a combination of nonlinear subunits. Estimating parameters of these models is made additionally more difficult by cortical neurons' trial-to-trial response variability. We develop a simple convolutional neural network method to estimate receptive field models for both simple and complex visual cortex cells from their responses to natural images. The model consists of a spatiotemporal filter, a parameterized rectifier unit (PReLU), and a two-dimensional Gaussian "map" of the receptive field envelope. A single model parameter determines the simple vs. complex nature of the receptive field, capturing complex cell responses as a summation of homogeneous subunits, and collapsing to a linear-nonlinear model for simple type cells. The convolutional method predicts simple and complex cell responses to natural image stimuli as well as grating tuning curves. The fitted models yield a continuum of values for the PReLU parameter across the sampled neurons, showing that the simple/complex nature of cells can vary in a continuous manner. We demonstrate that complex-like cells respond less reliably than simple-like cells. However, compensation for this unreliability with noise ceiling analysis reveals predictive performance for complex cells proportionately closer to that for simple cells. Most spatial receptive field structures are well fit by Gabor functions, whose parameters confirm well-known properties of cat A17/18 receptive fields.

Time-varying generalized linear models: characterizing and decoding neuronal dynamics in higher visual areas

To create a behaviorally relevant representation of the visual world, neurons in higher visual areas exhibit dynamic response changes to account for the time-varying interactions between external (e.g., visual input) and internal (e.g., reward value) factors. The resulting high-dimensional representational space poses challenges for precisely quantifying individual factors' contributions to the representation and readout of sensory information during a behavior. The widely used point process generalized linear model (GLM) approach provides a powerful framework for a quantitative description of neuronal processing as a function of various sensory and non-sensory inputs (encoding) as well as linking particular response components to particular behaviors (decoding), at the level of single trials and individual neurons. However, most existing variations of GLMs assume the neural systems to be time-invariant, making them inadequate for modeling nonstationary characteristics of neuronal sensitivity in higher visual areas. In this review, we summarize some of the existing GLM variations, with a focus on time-varying extensions. We highlight their applications to understanding neural representations in higher visual areas and decoding transient neuronal sensitivity as well as linking physiology to behavior through manipulation of model components. This time-varying class of statistical models provide valuable insights into the neural basis of various visual behaviors in higher visual areas and hold significant potential for uncovering the fundamental computational principles that govern neuronal processing underlying various behaviors in different regions of the brain.

Characterization of extracellular spike waveforms recorded in wallaby primary visual cortex

Extracellular recordings were made from 642 units in the primary visual cortex (V1) of a highly visual marsupial, the Tammar wallaby. The receptive field (RF) characteristics of the cells were objectively estimated using the non-linear input model (NIM), and these were correlated with spike shapes. We found that wallaby cortical units had 68% regular spiking (RS), 12% fast spiking (FS), 4% triphasic spiking (TS), 5% compound spiking (CS) and 11% positive spiking (PS). RS waveforms are most often associated with recordings from pyramidal or spiny stellate cell bodies, suggesting that recordings from these cell types dominate in the wallaby cortex. In wallaby, 70-80% of FS and RS cells had orientation selective RFs and had evenly distributed linear and nonlinear RFs. We found that 47% of wallaby PS units were non-orientation selective and they were dominated by linear RFs. Previous studies suggest that the PS units represent recordings from the axon terminals of non-orientation selective cells originating in the lateral geniculate nucleus (LGN). If this is also true in wallaby, as strongly suggested by their low response latencies and bursty spiking properties, the results suggest that significantly more neurons in wallaby LGN are already orientation selective. In wallaby, less than 10% of recorded spikes had triphasic (TS) or sluggish compound spiking (CS) waveforms. These units had a mixture of orientation selective and non-oriented properties, and their cellular origins remain difficult to classify.

Large-scale electrophysiology and deep learning reveal distorted neural signal dynamics after hearing loss

Listeners with hearing loss often struggle to understand speech in noise, even with a hearing aid. To better understand the auditory processing deficits that underlie this problem, we made large-scale brain recordings from gerbils, a common animal model for human hearing, while presenting a large database of speech and noise sounds. We first used manifold learning to identify the neural subspace in which speech is encoded and found that it is low-dimensional and that the dynamics within it are profoundly distorted by hearing loss. We then trained a deep neural network (DNN) to replicate the neural coding of speech with and without hearing loss and analyzed the underlying network dynamics. We found that hearing loss primarily impacts spectral processing, creating nonlinear distortions in cross-frequency interactions that result in a hypersensitivity to background noise that persists even after amplification with a hearing aid. Our results identify a new focus for efforts to design improved hearing aids and demonstrate the power of DNNs as a tool for the study of central brain structures.

SpikeSEE: An energy-efficient dynamic scenes processing framework for retinal prostheses

Intelligent and low-power retinal prostheses are highly demanded in this era, where wearable and implantable devices are used for numerous healthcare applications. In this paper, we propose an energy-efficient dynamic scenes processing framework (SpikeSEE) that combines a spike representation encoding technique and a bio-inspired spiking recurrent neural network (SRNN) model to achieve intelligent processing and extreme low-power computation for retinal prostheses. The spike representation encoding technique could interpret dynamic scenes with sparse spike trains, decreasing the data volume. The SRNN model, inspired by the human retina's special structure and spike processing method, is adopted to predict the response of ganglion cells to dynamic scenes. Experimental results show that the Pearson correlation coefficient of the proposed SRNN model achieves 0.93, which outperforms the state-of-the-art processing framework for retinal prostheses. Thanks to the spike representation and SRNN processing, the model can extract visual features in a multiplication-free fashion. The framework achieves 8 times power reduction compared with the convolutional recurrent neural network (CRNN) processing-based framework. Our proposed SpikeSEE predicts the response of ganglion cells more accurately with lower energy consumption, which alleviates the precision and power issues of retinal prostheses and provides a potential solution for wearable or implantable prostheses.

A convolutional neural network provides a generalizable model of natural sound coding by neural populations in auditory cortex

Convolutional neural networks (CNNs) can provide powerful and flexible models of neural sensory processing. However, the utility of CNNs in studying the auditory system has been limited by their requirement for large datasets and the complex response properties of single auditory neurons. To address these limitations, we developed a population encoding model: a CNN that simultaneously predicts activity of several hundred neurons recorded during presentation of a large set of natural sounds. This approach defines a shared spectro-temporal space and pools statistical power across neurons. Population models of varying architecture performed consistently and substantially better than traditional linear-nonlinear models on data from primary and non-primary auditory cortex. Moreover, population models were highly generalizable. The output layer of a model pre-trained on one population of neurons could be fit to data from novel single units, achieving performance equivalent to that of neurons in the original fit data. This ability to generalize suggests that population encoding models capture a complete representational space across neurons in an auditory cortical field.

Artificial intelligence techniques for retinal prostheses: a comprehensive review and future direction

Objective. Retinal prostheses are promising devices to restore vision for patients with severe age-related macular degeneration or retinitis pigmentosa disease. The visual processing mechanism embodied in retinal prostheses play an important role in the restoration effect. Its performance depends on our understanding of the retina's working mechanism and the evolvement of computer vision models. Recently, remarkable progress has been made in the field of processing algorithm for retinal prostheses where the new discovery of the retina's working principle and state-of-the-arts computer vision models are combined together.Approach. We investigated the related research on artificial intelligence techniques for retinal prostheses. The processing algorithm in these studies could be attributed to three types: computer vision-related methods, biophysical models, and deep learning models.Main results. In this review, we first illustrate the structure and function of the normal and degenerated retina, then demonstrate the vision rehabilitation mechanism of three representative retinal prostheses. It is necessary to summarize the computational frameworks abstracted from the normal retina. In addition, the development and feature of three types of different processing algorithms are summarized. Finally, we analyze the bottleneck in existing algorithms and propose our prospect about the future directions to improve the restoration effect.Significance. This review systematically summarizes existing processing models for predicting the response of the retina to external stimuli. What's more, the suggestions for future direction may inspire researchers in this field to design better algorithms for retinal prostheses.

Orientation pinwheels in primary visual cortex of a highly visual marsupial

Primary visual cortices in many mammalian species exhibit modular and periodic orientation preference maps arranged in pinwheel-like layouts. The role of inherited traits as opposed to environmental influences in determining this organization remains unclear. Here, we characterize the cortical organization of an Australian marsupial, revealing pinwheel organization resembling that of eutherian carnivores and primates but distinctly different from the simpler salt-and-pepper arrangement of eutherian rodents and rabbits. The divergence of marsupials from eutherians 160 million years ago and the later emergence of rodents and rabbits suggest that the salt-and-pepper structure is not the primitive ancestral form. Rather, the genetic code that enables complex pinwheel formation is likely widespread, perhaps extending back to the common therian ancestors of modern mammals.

Model-based characterization of the selectivity of neurons in primary visual cortex

Statistical models are increasingly being used to understand the complexity of stimulus selectivity in primary visual cortex (V1) in the context of complex time-varying stimuli, replacing averaging responses to simple parametric stimuli. Although such models often can more accurately reflect the computations performed by V1 neurons in more natural visual environments, they do not by themselves provide insight into V1 neural selectivity to basic stimulus features such as receptive field size, spatial frequency tuning, and phase invariance. Here, we present a battery of analyses that can be directly applied to encoding models to link complex encoding models to more interpretable aspects of stimulus selectivity. We apply this battery to nonlinear models of V1 neurons recorded in awake macaque during random bar stimuli. In linking model properties to more classical measurements, we demonstrate several novel aspects of V1 selectivity not available to simpler experimental measurements. For example, this approach reveals that individual spatiotemporal elements of the V1 models often have a smaller spatial scale than the neuron as a whole, resulting in nontrivial tuning to spatial frequencies. In addition, we propose measures of nonlinear integration that suggest that classical classifications of V1 neurons into simple versus complex cells will be spatial-frequency dependent. In total, rather than obfuscate classical characterizations of V1 neurons, model-based characterizations offer a means to more fully understand their selectivity, and link their classical tuning properties to their roles in more complex, natural, visual processing.NEW & NOTEWORTHY Visual neurons are increasingly being studied with more complex, natural visual stimuli, and increasingly complex models are necessary to characterize their response properties. Here, we describe a battery of analyses that relate these more complex models to classical characterizations. Using such model-based characterizations of V1 neurons furthermore yields several new insights into V1 processing not possible to capture in more classical means to measure their visual selectivity.

How Stimulus Statistics Affect the Receptive Fields of Cells in Primary Visual Cortex

We studied the changes that neuronal receptive field (RF) models undergo when the statistics of the stimulus are changed from those of white Gaussian noise (WGN) to those of natural scenes (NSs), by fitting the models to multielectrode data recorded from primary visual cortex (V1) of female cats. This allowed the estimation of both a cascade of linear filters on the stimulus, as well as the static nonlinearities that map the output of the filters to the neuronal spike rates. We found that cells respond differently to these two classes of stimuli, with mostly higher spike rates and shorter response latencies to NSs than to WGN. The most striking finding was that NSs resulted in RFs that had additional uncovered filters compared with WGN. This finding was not an artifact of the higher spike rates observed for NSs relative to WGN, but rather was related to a change in coding. Our results reveal a greater extent of nonlinear processing in V1 neurons when stimulated using NSs compared with WGN. Our findings indicate the existence of nonlinear mechanisms that endow V1 neurons with context-dependent transmission of visual information.SIGNIFICANCE STATEMENT This study addresses a fundamental question about the concept of the receptive field (RF): does the encoding of information depend on the context or statistical regularities of the stimulus type? We applied state-of-the-art RF modeling techniques to data collected from multielectrode recordings from cat visual cortex in response to two statistically distinct stimulus types: white Gaussian noise and natural scenes. We find significant differences between the RFs that emerge from our data-driven modeling. Natural scenes result in far more complex RFs that combine multiple features in the visual input. Our findings reveal that different regimes or modes of operation are at work in visual cortical processing depending on the information present in the visual input. The complexity of V1 neural coding appears to be dependent on the complexity of the stimulus. We believe this new finding will have interesting implications for our understanding of the efficient transmission of information in sensory systems, which is an integral assumption of many computational theories (e.g., efficient and predictive coding of sensory processing in the brain).

Retinal Encoding of Natural Scenes

An ultimate goal in retina science is to understand how the neural circuit of the retina processes natural visual scenes. Yet most studies in laboratories have long been performed with simple, artificial visual stimuli such as full-field illumination, spots of light, or gratings. The underlying assumption is that the features of the retina thus identified carry over to the more complex scenario of natural scenes. As the application of corresponding natural settings is becoming more commonplace in experimental investigations, this assumption is being put to the test and opportunities arise to discover processing features that are triggered by specific aspects of natural scenes. Here, we review how natural stimuli have been used to probe, refine, and complement knowledge accumulated under simplified stimuli, and we discuss challenges and opportunities along the way toward a comprehensive understanding of the encoding of natural scenes. Expected final online publication date for the Annual Review of Vision Science, Volume 8 is September 2022. Please see http://www.annualreviews.org/page/journal/pubdates for revised estimates.

Neural System Identification With Spike-Triggered Non-Negative Matrix Factorization

Neuronal circuits formed in the brain are complex with intricate connection patterns. Such complexity is also observed in the retina with a relatively simple neuronal circuit. A retinal ganglion cell (GC) receives excitatory inputs from neurons in previous layers as driving forces to fire spikes. Analytical methods are required to decipher these components in a systematic manner. Recently a method called spike-triggered non-negative matrix factorization (STNMF) has been proposed for this purpose. In this study, we extend the scope of the STNMF method. By using retinal GCs as a model system, we show that STNMF can detect various computational properties of upstream bipolar cells (BCs), including spatial receptive field, temporal filter, and transfer nonlinearity. In addition, we recover synaptic connection strengths from the weight matrix of STNMF. Furthermore, we show that STNMF can separate spikes of a GC into a few subsets of spikes, where each subset is contributed by one presynaptic BC. Taken together, these results corroborate that STNMF is a useful method for deciphering the structure of neuronal circuits.

NeuroSEE: A Neuromorphic Energy Efficient Processing Framework for Visual Prostheses

Visual prostheses with both comprehensive visual signal processing capability and energy efficiency are becoming increasingly demanded in the age of intelligent personal healthcare, particularly with the rise of wearable and implantable devices. To address this trend, we propose NeuroSEE, a neuromorphic energy-efficient processing framework that combines a spike representation encoding technique and a bio-inspired processing method. This framework first utilizes sparse spike trains to represent visual information, and then a bio-inspired spiking neural network (SNN) is adopted to process the spike trains. The SNN model makes use of an IF neuron with multiple spikefiring rates to decrease the energy consumption without compensating for prediction performance. The experimental results indicate that when predicting the response of the primary visual cortex, the framework achieves a state-ofthe- art Pearson correlation coefficient performance. Spikebased recording and processing methods simplify the storage and transmission of redundant scene information and complex calculation processes. It could reduce power consumption by 15 times compared with the existing Convolutional neural network (CNN) processing framework. The proposed NeuroSEE framework predicts the response of the primary visual cortex in an energy efficient manner, making it a powerful tool for visual prostheses.

Simple model for encoding natural images by retinal ganglion cells with nonlinear spatial integration

A central goal in sensory neuroscience is to understand the neuronal signal processing involved in the encoding of natural stimuli. A critical step towards this goal is the development of successful computational encoding models. For ganglion cells in the vertebrate retina, the development of satisfactory models for responses to natural visual scenes is an ongoing challenge. Standard models typically apply linear integration of visual stimuli over space, yet many ganglion cells are known to show nonlinear spatial integration, in particular when stimulated with contrast-reversing gratings. We here study the influence of spatial nonlinearities in the encoding of natural images by ganglion cells, using multielectrode-array recordings from isolated salamander and mouse retinas. We assess how responses to natural images depend on first- and second-order statistics of spatial patterns inside the receptive field. This leads us to a simple extension of current standard ganglion cell models. We show that taking not only the weighted average of light intensity inside the receptive field into account but also its variance over space can partly account for nonlinear integration and substantially improve response predictions of responses to novel images. For salamander ganglion cells, we find that response predictions for cell classes with large receptive fields profit most from including spatial contrast information. Finally, we demonstrate how this model framework can be used to assess the spatial scale of nonlinear integration. Our results underscore that nonlinear spatial stimulus integration translates to stimulation with natural images. Furthermore, the introduced model framework provides a simple, yet powerful extension of standard models and may serve as a benchmark for the development of more detailed models of the nonlinear structure of receptive fields.

For understanding how sensory systems operate in the natural environment, an important goal is to develop models that capture neuronal responses to natural stimuli. For retinal ganglion cells, which connect the eye to the brain, current standard models often fail to capture responses to natural visual scenes. This shortcoming is at least partly rooted in the fact that ganglion cells may combine visual signals over space in a nonlinear fashion. We here show that a simple model, which not only considers the average light intensity inside a cell’s receptive field but also the variance of light intensity over space, can partly account for these nonlinearities and thereby improve current standard models. This provides an easy-to-obtain benchmark for modeling ganglion cell responses to natural images.

Revealing Fine Structures of the Retinal Receptive Field by Deep-Learning Networks

Deep convolutional neural networks (CNNs) have demonstrated impressive performance on many visual tasks. Recently, they became useful models for the visual system in neuroscience. However, it is still not clear what is learned by CNNs in terms of neuronal circuits. When a deep CNN with many layers is used for the visual system, it is not easy to compare the structure components of CNNs with possible neuroscience underpinnings due to highly complex circuits from the retina to the higher visual cortex. Here, we address this issue by focusing on single retinal ganglion cells with biophysical models and recording data from animals. By training CNNs with white noise images to predict neuronal responses, we found that fine structures of the retinal receptive field can be revealed. Specifically, convolutional filters learned are resembling biological components of the retinal circuit. This suggests that a CNN learning from one single retinal cell reveals a minimal neural network carried out in this cell. Furthermore, when CNNs learned from different cells are transferred between cells, there is a diversity of transfer learning performance, which indicates that CNNs are cell specific. Moreover, when CNNs are transferred between different types of input images, here white noise versus natural images, transfer learning shows a good performance, which implies that CNNs indeed capture the full computational ability of a single retinal cell for different inputs. Taken together, these results suggest that CNNs could be used to reveal structure components of neuronal circuits, and provide a powerful model for neural system identification.

Complexity and diversity in sparse code priors improve receptive field characterization of Macaque V1 neurons

System identification techniques-projection pursuit regression models (PPRs) and convolutional neural networks (CNNs)-provide state-of-the-art performance in predicting visual cortical neurons' responses to arbitrary input stimuli. However, the constituent kernels recovered by these methods are often noisy and lack coherent structure, making it difficult to understand the underlying component features of a neuron's receptive field. In this paper, we show that using a dictionary of diverse kernels with complex shapes learned from natural scenes based on efficient coding theory, as the front-end for PPRs and CNNs can improve their performance in neuronal response prediction as well as algorithmic data efficiency and convergence speed. Extensive experimental results also indicate that these sparse-code kernels provide important information on the component features of a neuron's receptive field. In addition, we find that models with the complex-shaped sparse code front-end are significantly better than models with a standard orientation-selective Gabor filter front-end for modeling V1 neurons that have been found to exhibit complex pattern selectivity. We show that the relative performance difference due to these two front-ends can be used to produce a sensitive metric for detecting complex selectivity in V1 neurons.

Parallel and Recurrent Cascade Models as a Unifying Force for Understanding Sub-cellular Computation

Neurons are very complicated computational devices, incorporating numerous non-linear processes, particularly in their dendrites. Biophysical models capture these processes directly by explicitly modelling physiological variables, such as ion channels, current flow, membrane capacitance, etc. However, another option for capturing the complexities of real neural computation is to use cascade models, which treat individual neurons as a cascade of linear and non-linear operations, akin to a multi-layer artificial neural network. Recent research has shown that cascade models can capture single-cell computation well, but there are still a number of sub-cellular, regenerative dendritic phenomena that they cannot capture, such as the interaction between sodium, calcium, and NMDA spikes in different compartments. Here, we propose that it is possible to capture these additional phenomena using parallel, recurrent cascade models, wherein an individual neuron is modelled as a cascade of parallel linear and non-linear operations that can be connected recurrently, akin to a multi-layer, recurrent, artificial neural network. Given their tractable mathematical structure, we show that neuron models expressed in terms of parallel recurrent cascades can themselves be integrated into multi-layered artificial neural networks and trained to perform complex tasks. We go on to discuss potential implications and uses of these models for artificial intelligence. Overall, we argue that parallel, recurrent cascade models provide an important, unifying tool for capturing single-cell computation and exploring the algorithmic implications of physiological phenomena.

Decision-related feedback in visual cortex lacks spatial selectivity

Feedback in the brain is thought to convey contextual information that underlies our flexibility to perform different tasks. Empirical and computational work on the visual system suggests this is achieved by targeting task-relevant neuronal subpopulations. We combine two tasks, each resulting in selective modulation by feedback, to test whether the feedback reflected the combination of both selectivities. We used visual feature-discrimination specified at one of two possible locations and uncoupled the decision formation from motor plans to report it, while recording in macaque mid-level visual areas. Here we show that although the behavior is spatially selective, using only task-relevant information, modulation by decision-related feedback is spatially unselective. Population responses reveal similar stimulus-choice alignments irrespective of stimulus relevance. The results suggest a common mechanism across tasks, independent of the spatial selectivity these tasks demand. This may reflect biological constraints and facilitate generalization across tasks. Our findings also support a previously hypothesized link between feature-based attention and decision-related activity.

Feedback modulates visual neurons, thought to help achieve flexible task performance. Here, the authors show decision-related feedback is not only relayed to task-relevant neurons, suggesting a broader mechanism and supporting a previously hypothesized link to feature-based attention.

Learning divisive normalization in primary visual cortex

Divisive normalization (DN) is a prominent computational building block in the brain that has been proposed as a canonical cortical operation. Numerous experimental studies have verified its importance for capturing nonlinear neural response properties to simple, artificial stimuli, and computational studies suggest that DN is also an important component for processing natural stimuli. However, we lack quantitative models of DN that are directly informed by measurements of spiking responses in the brain and applicable to arbitrary stimuli. Here, we propose a DN model that is applicable to arbitrary input images. We test its ability to predict how neurons in macaque primary visual cortex (V1) respond to natural images, with a focus on nonlinear response properties within the classical receptive field. Our model consists of one layer of subunits followed by learned orientation-specific DN. It outperforms linear-nonlinear and wavelet-based feature representations and makes a significant step towards the performance of state-of-the-art convolutional neural network (CNN) models. Unlike deep CNNs, our compact DN model offers a direct interpretation of the nature of normalization. By inspecting the learned normalization pool of our model, we gained insights into a long-standing question about the tuning properties of DN that update the current textbook description: we found that within the receptive field oriented features were normalized preferentially by features with similar orientation rather than non-specifically as currently assumed.

State-space optimal feedback control of optogenetically driven neural activity

Objective.The rapid acceleration of tools for recording neuronal populations and targeted optogenetic manipulation has enabled real-time, feedback control of neuronal circuits in the brain. Continuously-graded control of measured neuronal activity poses a wide range of technical challenges, which we address through a combination of optogenetic stimulation and a state-space optimal control framework implemented in the thalamocortical circuit of the awake mouse.Approach.Closed-loop optogenetic control of neurons was performed in real-time via stimulation of channelrhodopsin-2 expressed in the somatosensory thalamus of the head-fixed mouse. A state-space linear dynamical system model structure was used to approximate the light-to-spiking input-output relationship in both single-neuron as well as multi-neuron scenarios when recording from multielectrode arrays. These models were utilized to design state feedback controller gains by way of linear quadratic optimal control and were also used online for estimation of state feedback, where a parameter-adaptive Kalman filter provided robustness to model-mismatch.Main results.This model-based control scheme proved effective for feedback control of single-neuron firing rate in the thalamus of awake animals. Notably, the graded optical actuation utilized here did not synchronize simultaneously recorded neurons, but heterogeneity across the neuronal population resulted in a varied response to stimulation. Simulated multi-output feedback control provided better control of a heterogeneous population and demonstrated how the approach generalizes beyond single-neuron applications.Significance.To our knowledge, this work represents the first experimental application of state space model-based feedback control for optogenetic stimulation. In combination with linear quadratic optimal control, the approaches laid out and tested here should generalize to future problems involving the control of highly complex neural circuits. More generally, feedback control of neuronal circuits opens the door to adaptively interacting with the dynamics underlying sensory, motor, and cognitive signaling, enabling a deeper understanding of circuit function and ultimately the control of function in the face of injury or disease.

Thalamic state influences timing precision in the thalamocortical circuit

Sensory signals from the outside world are transduced at the periphery, passing through thalamus before reaching cortex, ultimately giving rise to the sensory representations that enable us to perceive the world. The thalamocortical circuit is particularly sensitive to the temporal precision of thalamic spiking due to highly convergent synaptic connectivity. Thalamic neurons can exhibit burst and tonic modes of firing that strongly influence timing within the thalamus. The impact of these changes in thalamic state on sensory encoding in the cortex, however, remains unclear. Here, we investigated the role of thalamic state on timing in the thalamocortical circuit of the vibrissa pathway in the anesthetized rat. We optogenetically hyperpolarized thalamus while recording single unit activity in both thalamus and cortex. Tonic spike-triggered analysis revealed temporally precise thalamic spiking that was locked to weak white-noise sensory stimuli, whereas thalamic burst spiking was associated with a loss in stimulus-locked temporal precision. These thalamic state-dependent changes propagated to cortex such that the cortical timing precision was diminished during the hyperpolarized (burst biased) thalamic state. Although still sensory driven, the cortical neurons became significantly less precisely locked to the weak white-noise stimulus. The results here suggests a state-dependent differential regulation of spike timing precision in the thalamus that could gate what signals are ultimately propagated to cortex.NEW & NOTEWORTHY The majority of sensory signals are transmitted through the thalamus. There is growing evidence of complex thalamic gating through coordinated firing modes that have a strong impact on cortical sensory representations. Optogenetic hyperpolarization of thalamus pushed it into burst firing that disrupted precise time-locked sensory signaling, with a direct impact on the downstream cortical encoding, setting the stage for a timing-based thalamic gate of sensory signaling.

Analysis of extracellular spike waveforms and associated receptive fields of neurons in cat primary visual cortex

KEY POINTS

Extracellular spikes recorded in the visual cortex (Area 17/18, V1) are commonly classified into either regular-spiking (RS) or fast-spiking (FS). Using multi-electrode arrays positioned in cat V1 and a broadband stimulus, we show that there is also a distinct class with positive-spiking (PS) waveforms. PS units were associated mainly with non-oriented receptive fields while RS and FS units had orientation-selective receptive fields. We suggest that PS units are recordings of axons originating from the thalamus. This conclusion was reinforced by our finding that we could record PS units after cortical silencing, but not record RS and FS units. The importance of our findings is that we were able to correlate spike shapes with receptive field characteristics with high precision using multi-electrode extracellular recording techniques. This allows considerable increases in the amount of information that can be extracted from future cortical experiments.

ABSTRACT

Extracellular spike waveforms from recordings in the visual cortex have been classified into either regular-spiking (RS) or fast-spiking (FS) units. While both these types of spike waveforms are negative-dominant, we show that there are also distinct classes of spike waveforms in visual Area 17/18 (V1) of anaesthetised cats with positive-dominant waveforms, which are not regularly reported. The spatial receptive fields (RFs) of these different spike waveform types were estimated, which objectively revealed the existence of oriented and non-oriented RFs. We found that units with positive-dominant spikes, which have been associated with recordings from axons in the literature, had mostly non-oriented RFs (84%), which are similar to the centre-surround RFs observed in the dorsal lateral geniculate nucleus (dLGN). Thus, we hypothesise that these positive-dominant waveforms may be recordings from dLGN afferents. We recorded from V1 before and after the application of muscimol (a cortical silencer) and found that the positive-dominant spikes (PS) remained while the RS and FS cells did not. We also noted that the PS units had spiking characteristics normally associated with dLGN units (i.e. higher response spike rates, lower response latencies and higher proportion of burst spikes). Our findings show quantitatively that it is possible to correlate the RF properties of cortical neurons with particular spike waveforms. This has implications for how extracellular recordings should be interpreted and complex experiments can now be contemplated that would have been very challenging previously, such as assessing the feedforward connectivity between brain areas in the same location of cortical tissue.

Learning receptive field properties of complex cells in V1

There are two distinct classes of cells in the primary visual cortex (V1): simple cells and complex cells. One defining feature of complex cells is their spatial phase invariance; they respond strongly to oriented grating stimuli with a preferred orientation but with a wide range of spatial phases. A classical model of complete spatial phase invariance in complex cells is the energy model, in which the responses are the sum of the squared outputs of two linear spatially phase-shifted filters. However, recent experimental studies have shown that complex cells have a diverse range of spatial phase invariance and only a subset can be characterized by the energy model. While several models have been proposed to explain how complex cells could learn to be selective to orientation but invariant to spatial phase, most existing models overlook many biologically important details. We propose a biologically plausible model for complex cells that learns to pool inputs from simple cells based on the presentation of natural scene stimuli. The model is a three-layer network with rate-based neurons that describes the activities of LGN cells (layer 1), V1 simple cells (layer 2), and V1 complex cells (layer 3). The first two layers implement a recently proposed simple cell model that is biologically plausible and accounts for many experimental phenomena. The neural dynamics of the complex cells is modeled as the integration of simple cells inputs along with response normalization. Connections between LGN and simple cells are learned using Hebbian and anti-Hebbian plasticity. Connections between simple and complex cells are learned using a modified version of the Bienenstock, Cooper, and Munro (BCM) rule. Our results demonstrate that the learning rule can describe a diversity of complex cells, similar to those observed experimentally.

Many cortical functions originate from the learning ability of the brain. How the properties of cortical cells are learned is vital for understanding how the brain works. There are many models that explain how V1 simple cells can be learned. However, how V1 complex cells are learned still remains unclear. In this paper, we propose a model of learning in complex cells based on the Bienenstock, Cooper, and Munro (BCM) rule. We demonstrate that properties of receptive fields of complex cells can be learned using this biologically plausible learning rule. Quantitative comparisons between the model and experimental data are performed. Results show that model complex cells can account for the diversity of complex cells found in experimental studies. In summary, this study provides a plausible explanation for how complex cells can be learned using biologically plausible plasticity mechanisms. Our findings help us to better understand biological vision processing and provide us with insights into the general signal processing principles that the visual cortex employs to process visual information.

Linear-nonlinear cascades capture synaptic dynamics

Short-term synaptic dynamics differ markedly across connections and strongly regulate how action potentials communicate information. To model the range of synaptic dynamics observed in experiments, we have developed a flexible mathematical framework based on a linear-nonlinear operation. This model can capture various experimentally observed features of synaptic dynamics and different types of heteroskedasticity. Despite its conceptual simplicity, we show that it is more adaptable than previous models. Combined with a standard maximum likelihood approach, synaptic dynamics can be accurately and efficiently characterized using naturalistic stimulation patterns. These results make explicit that synaptic processing bears algorithmic similarities with information processing in convolutional neural networks.

Atoh7-independent specification of retinal ganglion cell identity

Retinal ganglion cells (RGCs) relay visual information from the eye to the brain. RGCs are the first cell type generated during retinal neurogenesis. Loss of function of the transcription factor Atoh7, expressed in multipotent early neurogenic retinal progenitors leads to a selective and essentially complete loss of RGCs. Therefore, Atoh7 is considered essential for conferring competence on progenitors to generate RGCs. Despite the importance of Atoh7 in RGC specification, we find that inhibiting apoptosis in Atoh7-deficient mice by loss of function of Bax only modestly reduces RGC numbers. Single-cell RNA sequencing of Atoh7;Bax-deficient retinas shows that RGC differentiation is delayed but that the gene expression profile of RGC precursors is grossly normal. Atoh7;Bax-deficient RGCs eventually mature, fire action potentials, and incorporate into retinal circuitry but exhibit severe axonal guidance defects. This study reveals an essential role for Atoh7 in RGC survival and demonstrates Atoh7-dependent and Atoh7-independent mechanisms for RGC specification.

Predicting synchronous firing of large neural populations from sequential recordings

A major goal in neuroscience is to understand how populations of neurons code for stimuli or actions. While the number of neurons that can be recorded simultaneously is increasing at a fast pace, in most cases these recordings cannot access a complete population: some neurons that carry relevant information remain unrecorded. In particular, it is hard to simultaneously record all the neurons of the same type in a given area. Recent progress have made possible to profile each recorded neuron in a given area thanks to genetic and physiological tools, and to pool together recordings from neurons of the same type across different experimental sessions. However, it is unclear how to infer the activity of a full population of neurons of the same type from these sequential recordings. Neural networks exhibit collective behaviour, e.g. noise correlations and synchronous activity, that are not directly captured by a conditionally-independent model that would just put together the spike trains from sequential recordings. Here we show that we can infer the activity of a full population of retina ganglion cells from sequential recordings, using a novel method based on copula distributions and maximum entropy modeling. From just the spiking response of each ganglion cell to a repeated stimulus, and a few pairwise recordings, we could predict the noise correlations using copulas, and then the full activity of a large population of ganglion cells of the same type using maximum entropy modeling. Remarkably, we could generalize to predict the population responses to different stimuli with similar light conditions and even to different experiments. We could therefore use our method to construct a very large population merging cells' responses from different experiments. We predicted that synchronous activity in ganglion cell populations saturates only for patches larger than 1.5mm in radius, beyond what is today experimentally accessible.

Computational challenges and opportunities for a bi-directional artificial retina

A future artificial retina that can restore high acuity vision in blind people will rely on the capability to both read (observe) and write (control) the spiking activity of neurons using an adaptive, bi-directional and high-resolution device. Although current research is focused on overcoming the technical challenges of building and implanting such a device, exploiting its capabilities to achieve more acute visual perception will also require substantial computational advances. Using high-density large-scale recording and stimulation in the primate retina with an ex vivo multi-electrode array lab prototype, we frame several of the major computational problems, and describe current progress and future opportunities in solving them. First, we identify cell types and locations from spontaneous activity in the blind retina, and then efficiently estimate their visual response properties by using a low-dimensional manifold of inter-retina variability learned from a large experimental dataset. Second, we estimate retinal responses to a large collection of relevant electrical stimuli by passing current patterns through an electrode array, spike sorting the resulting recordings and using the results to develop a model of evoked responses. Third, we reproduce the desired responses for a given visual target by temporally dithering a diverse collection of electrical stimuli within the integration time of the visual system. Together, these novel approaches may substantially enhance artificial vision in a next-generation device.

Capturing Multiple Timescales of Adaptation to Second-Order Statistics With Generalized Linear Models: Gain Scaling and Fractional Differentiation

Single neurons can dynamically change the gain of their spiking responses to take into account shifts in stimulus variance. Moreover, gain adaptation can occur across multiple timescales. Here, we examine the ability of a simple statistical model of spike trains, the generalized linear model (GLM), to account for these adaptive effects. The GLM describes spiking as a Poisson process whose rate depends on a linear combination of the stimulus and recent spike history. The GLM successfully replicates gain scaling observed in Hodgkin-Huxley simulations of cortical neurons that occurs when the ratio of spike-generating potassium and sodium conductances approaches one. Gain scaling in the GLM depends on the length and shape of the spike history filter. Additionally, the GLM captures adaptation that occurs over multiple timescales as a fractional derivative of the stimulus envelope, which has been observed in neurons that include long timescale afterhyperpolarization conductances. Fractional differentiation in GLMs requires long spike history that span several seconds. Together, these results demonstrate that the GLM provides a tractable statistical approach for examining single-neuron adaptive computations in response to changes in stimulus variance.

Ignoring correlated activity causes a failure of retinal population codes

From starlight to sunlight, adaptation alters retinal output, changing both the signal and noise among populations of retinal ganglion cells (RGCs). Here we determine how these light level-dependent changes impact decoding of retinal output, testing the importance of accounting for RGC noise correlations to optimally read out retinal activity. We find that at moonlight conditions, correlated noise is greater and assuming independent noise severely diminishes decoding performance. In fact, assuming independence among a local population of RGCs produces worse decoding than using a single RGC, demonstrating a failure of population codes when correlated noise is substantial and ignored. We generalize these results with a simple model to determine what conditions dictate this failure of population processing. This work elucidates the circumstances in which accounting for noise correlations is necessary to take advantage of population-level codes and shows that sensory adaptation can strongly impact decoding requirements on downstream brain areas.

To see during day and night, the retina adapts to a trillion-fold change in light intensity. The authors show that an accurate read-out of retinal signals over this intensity range requires that brain circuits account for changing noise correlations across populations of retinal neurons.

Binless Kernel Machine: Modeling Spike Train Transformation for Cognitive Neural Prostheses

Modeling spike train transformation among brain regions helps in designing a cognitive neural prosthesis that restores lost cognitive functions. Various methods analyze the nonlinear dynamic spike train transformation between two cortical areas with low computational eficiency. The application of a real-time neural prosthesis requires computational eficiency, performance stability, and better interpretation of the neural firing patterns that modulate target spike generation. We propose the binless kernel machine in the point-process framework to describe nonlinear dynamic spike train transformations. Our approach embeds the binless kernel to eficiently capture the feedforward dynamics of spike trains and maps the input spike timings into reproducing kernel Hilbert space (RKHS). An inhomogeneous Bernoulli process is designed to combine with a kernel logistic regression that operates on the binless kernel to generate an output spike train as a point process. Weights of the proposed model are estimated by maximizing the log likelihood of output spike trains in RKHS, which allows a global-optimal solution. To reduce computational complexity, we design a streaming-based clustering algorithm to extract typical and important spike train features. The cluster centers and their weights enable the visualization of the important input spike train patterns that motivate or inhibit output neuron firing. We test the proposed model on both synthetic data and real spike train data recorded from the dorsal premotor cortex and the primary motor cortex of a monkey performing a center-out task. Performances are evaluated by discrete-time rescaling Kolmogorov-Smirnov tests. Our model outperforms the existing methods with higher stability regardless of weight initialization and demonstrates higher eficiency in analyzing neural patterns from spike timing with less historical input (50%). Meanwhile, the typical spike train patterns selected according to weights are validated to encode output spike from the spike train of single-input neuron and the interaction of two input neurons.

Estimating and interpreting nonlinear receptive field of sensory neural responses with deep neural network models

Our understanding of nonlinear stimulus transformations by neural circuits is hindered by the lack of comprehensive yet interpretable computational modeling frameworks. Here, we propose a data-driven approach based on deep neural networks to directly model arbitrarily nonlinear stimulus-response mappings. Reformulating the exact function of a trained neural network as a collection of stimulus-dependent linear functions enables a locally linear receptive field interpretation of the neural network. Predicting the neural responses recorded invasively from the auditory cortex of neurosurgical patients as they listened to speech, this approach significantly improves the prediction accuracy of auditory cortical responses, particularly in nonprimary areas. Moreover, interpreting the functions learned by neural networks uncovered three distinct types of nonlinear transformations of speech that varied considerably from primary to nonprimary auditory regions. The ability of this framework to capture arbitrary stimulus-response mappings while maintaining model interpretability leads to a better understanding of cortical processing of sensory signals.

Modeling the Short-Term Dynamics of in Vivo Excitatory Spike Transmission

Information transmission in neural networks is influenced by both short-term synaptic plasticity (STP) as well as nonsynaptic factors, such as after-hyperpolarization currents and changes in excitability. Although these effects have been widely characterized in vitro using intracellular recordings, how they interact in vivo is unclear. Here, we develop a statistical model of the short-term dynamics of spike transmission that aims to disentangle the contributions of synaptic and nonsynaptic effects based only on observed presynaptic and postsynaptic spiking. The model includes a dynamic functional connection with short-term plasticity as well as effects due to the recent history of postsynaptic spiking and slow changes in postsynaptic excitability. Using paired spike recordings, we find that the model accurately describes the short-term dynamics of in vivo spike transmission at a diverse set of identified and putative excitatory synapses, including a pair of connected neurons within thalamus in mouse, a thalamocortical connection in a female rabbit, and an auditory brainstem synapse in a female gerbil. We illustrate the utility of this modeling approach by showing how the spike transmission patterns captured by the model may be sufficient to account for stimulus-dependent differences in spike transmission in the auditory brainstem (endbulb of Held). Finally, we apply this model to large-scale multielectrode recordings to illustrate how such an approach has the potential to reveal cell type-specific differences in spike transmission in vivo Although STP parameters estimated from ongoing presynaptic and postsynaptic spiking are highly uncertain, our results are partially consistent with previous intracellular observations in these synapses.SIGNIFICANCE STATEMENT Although synaptic dynamics have been extensively studied and modeled using intracellular recordings of postsynaptic currents and potentials, inferring synaptic effects from extracellular spiking is challenging. Whether or not a synaptic current contributes to postsynaptic spiking depends not only on the amplitude of the current, but also on many other factors, including the activity of other, typically unobserved, synapses, the overall excitability of the postsynaptic neuron, and how recently the postsynaptic neuron has spiked. Here, we developed a model that, using only observations of presynaptic and postsynaptic spiking, aims to describe the dynamics of in vivo spike transmission by modeling both short-term synaptic plasticity (STP) and nonsynaptic effects. This approach may provide a novel description of fast, structured changes in spike transmission.

Mechanisms of Feature Selectivity and Invariance in Primary Visual Cortex

Visual object identification requires both selectivity for specific visual features that are important to the object's identity and invariance to feature manipulations. For example, a hand can be shifted in position, rotated, or contracted but still be recognized as a hand. How are the competing requirements of selectivity and invariance built into the early stages of visual processing? Typically, cells in the primary visual cortex are classified as either simple or complex. They both show selectivity for edge-orientation but complex cells develop invariance to edge position within the receptive field (spatial phase). Using a data-driven model that extracts the spatial structures and nonlinearities associated with neuronal computation, we quantitatively describe the balance between selectivity and invariance in complex cells. Phase invariance is frequently partial, while invariance to orientation and spatial frequency are more extensive than expected. The invariance arises due to two independent factors: (1) the structure and number of filters and (2) the form of nonlinearities that act upon the filter outputs. Both vary more than previously considered, so primary visual cortex forms an elaborate set of generic feature sensitivities, providing the foundation for more sophisticated object processing.

The statistics of how natural images drive the responses of neurons

To model the responses of neurons in the early visual system, at least three basic components are required: a receptive field, a normalization term, and a specification of encoding noise. Here, we examine how the receptive field, the normalization factor, and the encoding noise affect the drive to model-neuron responses when stimulated with natural images. We show that when these components are modeled appropriately, the response drives elicited by natural stimuli are Gaussian-distributed and scale invariant, and very nearly maximize the sensitivity (d') for natural-image discrimination. We discuss the statistical models of natural stimuli that can account for these response statistics, and we show how some commonly used modeling practices may distort these results. Finally, we show that normalization can equalize important properties of neural response across different stimulus types. Specifically, narrowband (stimulus- and feature-specific) normalization causes model neurons to yield Gaussian response-drive statistics when stimulated with natural stimuli, 1/f noise stimuli, and white-noise stimuli. The current work makes recommendations for best practices and lays a foundation, grounded in the response statistics to natural stimuli, upon which to build principled models of more complex visual tasks.

Inference of nonlinear receptive field subunits with spike-triggered clustering

Responses of sensory neurons are often modeled using a weighted combination of rectified linear subunits. Since these subunits often cannot be measured directly, a flexible method is needed to infer their properties from the responses of downstream neurons. We present a method for maximum likelihood estimation of subunits by soft-clustering spike-triggered stimuli, and demonstrate its effectiveness in visual neurons. For parasol retinal ganglion cells in macaque retina, estimated subunits partitioned the receptive field into compact regions, likely representing aggregated bipolar cell inputs. Joint clustering revealed shared subunits between neighboring cells, producing a parsimonious population model. Closed-loop validation, using stimuli lying in the null space of the linear receptive field, revealed stronger nonlinearities in OFF cells than ON cells. Responses to natural images, jittered to emulate fixational eye movements, were accurately predicted by the subunit model. Finally, the generality of the approach was demonstrated in macaque V1 neurons.

Inferring synaptic inputs from spikes with a conductance-based neural encoding model

Descriptive statistical models of neural responses generally aim to characterize the mapping from stimuli to spike responses while ignoring biophysical details of the encoding process. Here, we introduce an alternative approach, the conductance-based encoding model (CBEM), which describes a mapping from stimuli to excitatory and inhibitory synaptic conductances governing the dynamics of sub-threshold membrane potential. Remarkably, we show that the CBEM can be fit to extracellular spike train data and then used to predict excitatory and inhibitory synaptic currents. We validate these predictions with intracellular recordings from macaque retinal ganglion cells. Moreover, we offer a novel quasi-biophysical interpretation of the Poisson generalized linear model (GLM) as a special case of the CBEM in which excitation and inhibition are perfectly balanced. This work forges a new link between statistical and biophysical models of neural encoding and sheds new light on the biophysical variables that underlie spiking in the early visual pathway.

Understanding the circuit basis of cognitive functions using mouse models

Understanding how cognitive functions arise from computations occurring in the brain requires the ability to measure and perturb neural activity while the relevant circuits are engaged for specific cognitive processes. Rapid technical advances have led to the development of new approaches to transiently activate and suppress neuronal activity as well as to record simultaneously from hundreds to thousands of neurons across multiple brain regions during behavior. To realize the full potential of these approaches for understanding cognition, however, it is critical that behavioral conditions and stimuli are effectively designed to engage the relevant brain networks. Here, we highlight recent innovations that enable this combined approach. In particular, we focus on how to design behavioral experiments that leverage the ever-growing arsenal of technologies for controlling and measuring neural activity in order to understand cognitive functions.

Postural Representations of the Hand in the Primate Sensorimotor Cortex

Manual dexterity requires proprioceptive feedback about the state of the hand. To date, study of the neural basis of proprioception in the cortex has focused primarily on reaching movements to the exclusion of hand-specific behaviors such as grasping. To fill this gap, we record both time-varying hand kinematics and neural activity evoked in somatosensory and motor cortices as monkeys grasp a variety of objects. We find that neurons in the somatosensory cortex, as well as in the motor cortex, preferentially track time-varying postures of multi-joint combinations spanning the entire hand. This contrasts with neural responses during reaching movements, which preferentially track time-varying movement kinematics of the arm, such as velocity and speed of the limb, rather than its time-varying postural configuration. These results suggest different representations of arm and hand movements suited to the different functional roles of these two effectors.

Multiple Timescales Account for Adaptive Responses across Sensory Cortices

Sensory systems encounter remarkably diverse stimuli in the external environment. Natural stimuli exhibit timescales and amplitudes of variation that span a wide range. Mechanisms of adaptation, a ubiquitous feature of sensory systems, allow for the accommodation of this range of scales. Are there common rules of adaptation across different sensory modalities? We measured the membrane potential responses of individual neurons in the visual, somatosensory, and auditory cortices of male and female mice to discrete, punctate stimuli delivered at a wide range of fixed and nonfixed frequencies. We find that the adaptive profile of the response is largely preserved across these three areas, exhibiting attenuation and responses to the cessation of stimulation, which are signatures of response to changes in stimulus statistics. We demonstrate that these adaptive responses can emerge from a simple model based on the integration of fixed filters operating over multiple time scales.SIGNIFICANCE STATEMENT Our recent sensations affect our current expectations and perceptions of the environment. Neural correlates of this process exist throughout the brain and are loosely termed adaptation. Adaptive processes have been described across sensory cortices, but direct comparisons of these processes have not been possible because paradigms have been tailored specifically for each modality. We developed a common stimulus set that was used to characterize adaptation in somatosensory, visual, and auditory cortex. We describe here the similarities and differences in adaptation across these cortical areas and demonstrate that adaptive responses may emerge from a set of static filters that operate over a broad range of timescales.

Design strategies for dynamic closed-loop optogenetic neurocontrol in vivo

OBJECTIVE

Controlling neural activity enables the possibility of manipulating sensory perception, cognitive processes, and body movement, in addition to providing a powerful framework for functionally disentangling the neural circuits that underlie these complex phenomena. Over the last decade, optogenetic stimulation has become an increasingly important and powerful tool for understanding neural circuit function, owing to the ability to target specific cell types and bidirectionally modulate neural activity. To date, most stimulation has been provided in open-loop or in an on/off closed-loop fashion, where previously-determined stimulation is triggered by an event. Here, we describe and demonstrate a design approach for precise optogenetic control of neuronal firing rate modulation using feedback to guide stimulation continuously.

APPROACH

Using the rodent somatosensory thalamus as an experimental testbed for realizing desired time-varying patterns of firing rate modulation, we utilized a moving average exponential filter to estimate firing rate online from single-unit spiking measured extracellularly. This estimate of instantaneous rate served as feedback for a proportional integral (PI) controller, which was designed during the experiment based on a linear-nonlinear Poisson (LNP) model of the neuronal response to light.

MAIN RESULTS

The LNP model fit during the experiment enabled robust closed-loop control, resulting in good tracking of sinusoidal and non-sinusoidal targets, and rejection of unmeasured disturbances. Closed-loop control also enabled manipulation of trial-to-trial variability.

SIGNIFICANCE

Because neuroscientists are faced with the challenge of dissecting the functions of circuit components, the ability to maintain control of a region of interest in spite of changes in ongoing neural activity will be important for disambiguating function within networks. Closed-loop stimulation strategies are ideal for control that is robust to such changes, and the employment of continuous feedback to adjust stimulation in real-time can improve the quality of data collected using optogenetic manipulation.

Data-Driven Approaches to Understanding Visual Neuron Activity

With modern neurophysiological methods able to record neural activity throughout the visual pathway in the context of arbitrarily complex visual stimulation, our understanding of visual system function is becoming limited by the available models of visual neurons that can be directly related to such data. Different forms of statistical models are now being used to probe the cellular and circuit mechanisms shaping neural activity, understand how neural selectivity to complex visual features is computed, and derive the ways in which neurons contribute to systems-level visual processing. However, models that are able to more accurately reproduce observed neural activity often defy simple interpretations. As a result, rather than being used solely to connect with existing theories of visual processing, statistical modeling will increasingly drive the evolution of more sophisticated theories.