Direction selectivity of cells in the cat's striate cortex: differences between bar and grating stimuli

The linearity and selectivity of neuronal responses in awake visual cortex

Neurons in primary visual cortex (V1) are frequently classified based on their response linearity: the extent to which their visual responses to drifting gratings resemble a linear replica of the stimulus. This classification is supported by the finding that response linearity is bimodally distributed across neurons in area V1 of anesthetized animals. However, recent studies suggest that such bimodal distribution may not reflect two neuronal types but a nonlinear relationship between the membrane potential and the spike output. A main limitation of these previous studies is that they measured response linearity in anesthetized animals, where the distance between the neuronal membrane potential and the spike threshold is artificially increased by anesthesia. Here, we measured V1 response linearity in the awake brain and its correlation with the neuronal spontaneous firing rate, which is related to the distance between membrane potential and threshold. Our results demonstrate that response linearity is bimodally distributed in awake V1 but that it is poorly correlated with spontaneous firing rate. In contrast, the spontaneous firing rate is best correlated to the response selectivity and response latency to stimuli.

Circuitry and the classification of simple and complex cells in V1

Based on a large-scale neural network model of striate cortex (V1), we present a simulation study of extra- and intracellular response modulations for drifting and contrast reversal grating stimuli. Specifically, we study the dependency of these modulations on the neural circuitry. We find that the frequently used ratio of the first harmonic to the mean response to classify simple and complex cells is highly insensitive to circuitry. Limited experimental sample size for the distribution of this measure makes it unsuitable for distinguishing whether the dichotomy of simple and complex cells originates from distinct LGN axon connectivity and/or local circuitry in V1. We show that a possible useful measure in this respect is the ratio of the intracellular second- to first-harmonic response for contrast reversal gratings. This measure is highly sensitive to neural circuitry and its distribution can be sampled with sufficient accuracy from a limited amount of experimental data. Further, the distribution of this measure is qualitatively similar to that of the subfield correlation coefficient, although it is more robust and easier to obtain experimentally.

Spatio-temporal requirements for direction selectivity in area 18 and PMLS complex cells

The spatio-temporal requirements for direction selectivity were studied in two extrastriate motion processing areas in the cat, area 18 and the posteromedial lateral suprasylvian cortex (PMLS). Direction, velocity and pixel size of random pixel arrays (RPA) were adjusted for each neuron and direction selectivity was measured as a function of step size and delay for a given optimal velocity. A subset of direction selective complex cells in area 18 was tuned to intermediate step size and delay combinations rather than the smoothest motion (band-pass cells). Other area 18 complex cells responded best to the smallest value of step size and delay (low-pass cells). Tuning varied with the pixel size of the RPA. Cells with tuning for smaller pixels favoured a preference for non-smooth motion. Area 18 cells with lower spatial resolution showed larger optimal and maximal step sizes. For a subset of the cells in area 18, we measured direction selectivity for extensive step-delay combinations, covering multiple velocities. Results showed that most cells were tuned to narrow range of step-delay combinations, and that the optimal step size was independent of temporal delay. Direction selective complex cells in PMLS were tuned to larger pixel sizes than those in area 18, although the distributions did overlap. In contrast to area 18, PMLS cells preferred the smoothest motion, irrespective of RPA pixel size.

Effect of interocular delay on disparity-selective v1 neurons: relationship to stereoacuity and the pulfrich effect

The temporal properties of disparity-sensitive neurons place important temporal constraints on stereo matching. We examined these constraints by measuring the responses of disparity-selective neurons in striate cortex of awake behaving monkeys to random-dot stereograms that contained interocular delays. Disparity selectivity was gradually abolished by increasing interocular delay (when the delay exceeds the integration time, the inputs from the 2 eyes become uncorrelated). The amplitude of the disparity-selective response was a Gaussian function of interocular delay, with a mean of 16 ms (+/-5 ms, SD). Psychophysical measures of stereoacuity, in both monkey and human observers, showed a closely similar dependency on time, suggesting that temporal integration in V1 neurons is what determines psychophysical matching constraints over time. There was a slight but consistent asymmetry in the neuronal responses, as if the optimum stimulus is one in which the right stimulus leads by about 4 ms. Because all recordings were made in the left hemisphere, this probably reflects nasotemporal differences in conduction times; psychophysical data are compatible with this interpretation. In only a few neurons (5/72), interocular delay caused a change in the preferred disparity. Such tilted disparity/delay profiles have been invoked previously to explain depth perception in the stroboscopic version of the Pulfrich effect (and other variants). However, the great majority of the neurons did not show tilted disparity/delay profiles. This suggests that either the activity of these neurons is ignored when viewing Pulfrich stimuli, or that current theories relating neuronal properties to perception in the Pulfrich effect need to be reevaluated.

A nonlinear model of the behavior of simple cells in visual cortex

Despite their structured receptive fields (RFs) and the strong linear components in their responses, most simple cells in mammalian visual cortex exhibit nonlinear behaviors. Besides the contrast-response function, nonlinearities are evident in various types of failure at superposition tasks, in the disagreement between direction indices computed from drifting and counterphase flickering gratings, in various forms of response suppression (including end- and side-stopping, spatial-frequency-specific inhibition and cross-orientation inhibition), in the advance of phase with increasing contrast, and in phase-insensitive and frequency-doubled responses to counterphase flickering gratings. These behaviors suggest that nonlinearities are involved in the operation of simple cells, but current models fail to explain them. A quantitative model is presented here that purports to describe basic and common principles of operation for all visual cortical cells. Simple cells are described as receiving afferents from multiple subunits that differ in their individual RFs and temporal impulse responses (TIRs). Subunits are independent and perform a spatial integration across their RFs followed by halfwave rectification and temporal convolution with their TIRs. This parallel operation yields a set of temporal functions representing each subunit's contribution to the membrane potential of the host cell, whose final form is given by the weighted sum of all subunits' contributions. By varying the number of subunits and their particular characteristics, different instances of the model are obtained each of which displays a different set of behaviors. Extensive simulation results are presented that illustrate how all of the reported nonlinear behaviors of simple cells arise from these multi-subunit organizations.

The derivation of direction selectivity in the striate cortex

In the central visual pathway of binocular animals, the property of directional selectivity (DS) is first exhibited in striate cortex. In this study, we sought to determine the neural circuitry underlying the transformation from non-DS neurons to DS cortical cells. In a well established model, DS receptive fields (RFs) are derived from the sum of two non-DS inputs with 90 degrees (quadrature) spatiotemporal phase differences. We explored possible input sources for this model, which include non-DS simple cells and lateral geniculate nucleus (LGN) neurons, by examination of spatiotemporal RFs of single cells and of pairs of cells. We find that distributions of non-DS simple RFs do not match the temporal predictions of the quadrature model because of a lack of long-latency responses. The long-latency inputs could potentially arise from lagged LGN afferents. However, analysis of cell pairs indicates that DS cells receive cortical input from non-DS simple cells for both short- and long-latency components, with temporal phase differences typically <90 degrees. Furthermore, the distribution of minimum phase differences needed to generate DS cells overlaps that exhibited by non-DS simple cells. Considered together, these results are consistent with a linear model whereby DS simple cells are formed from simple-cell inputs, with temporal phase differences often less than quadrature.

On the classification of simple and complex cells

In their pioneering studies of primary visual cortex, Hubel and Wiesel described the existence of two classes of cells, which they termed "simple" and "complex". The original classification scheme was based on a number of partly subjective tests of linear spatial summation. Later, investigators adopted an objective classification method based on the ratio between the amplitude of the first harmonic of the response and the mean spike rate (or the F(1)/F(0) ratio) when the neuron is stimulated with drifting sinusoidal gratings. This measure is bimodally distributed over the population and divides neurons into two classes that correspond closely to the classical definition by Hubel and Wiesel. Here we show that a simple rectification model can predict the observed bimodal distribution of F(1)/F(0) in primary visual cortex when the distributions of the intracellular response modulation and mean are unimodal. Thus, contrary to common belief, the bimodality of F(1)/F(0) does not necessarily imply the existence of two discrete cell classes. Furthermore, in reviewing the literature, we find no independent support for a simple/complex dichotomy. These results suggest that the existence of two distinct neural populations in primary visual cortex, and the associated hierarchical model of receptive field organization, need to be re-evaluated.

Responses of neurons in the cat posteromedial lateral suprasylvian cortex to moving texture patterns

The posteromedial lateral suprasylvian cortex represents a point of convergence between the geniculostriate and extrageniculostriate visual pathways. Given its purported role in motion analysis and the conflicting reports regarding the texture sensitivity of this area, we have investigated the response properties of cells in PMLS to moving texture patterns ("visual noise"). In contrast to previous reports, we have found that a large majority of cells (80.1%) responds to the motion of a texture pattern with sustained discharges. In general, responses to noise were more broadly tuned for direction compared to gratings; however, direction selectivity appeared more pronounced in response to noise. The majority of cells was selective for drift velocity of the noise pattern (mean optimal velocity: 26.7 degrees /s). Velocity tuning was comparable to that of its principal thalamic input, the lateral posterior pulvinar nucleus. In general, responsiveness of cells in the posteromedial lateral suprasylvian cortex increased with increasing texture element size, although some units were tuned to smaller element sizes than the largest presented. Finally, the magnitude of these noise responses was dependent on the area of the visual field stimulated. In general, a stimulus corresponding to roughly twice the size of the receptive field was required to elicit an equivalent half-maximal response to that for gratings. The results of this study indicate that the majority of cells in the posteromedial lateral suprasylvian cortex can be driven by the motion of a fine texture field, and highlight the importance of this area in motion analysis.

Direction selectivity and spatiotemporal separability in simple cortical cells

Simple cells in mammalian visual cortex are quasi-linear mechanisms whose behavior departs from true linearity in a very consistent manner. Empirical research on direction selectivity (DS) clearly illustrates these characteristics. A linear DS cell will be DS for all stimuli, whereas a linear non-DS cell will not be DS for any stimuli. However, many simple cells have opposite preferred directions for stimuli of reversed polarity, and some cells are DS for some stimuli (e.g., moving bars) but not for others (e.g., drifting gratings). Also, linear non-DS cells must have separable spatiotemporal receptive fields (RFs), and linear DS cells must have inseparable RFs. Yet many actual DS cells have separable RFs. Here we present a nonlinear model of simple-cell behavior that reproduces all of these empirical behaviors. The model is a variant of the current linear model, amended to include an interleaved nonlinearity (half-wave rectification) that allows it to mimic the (im)balance of push-pull mechanisms. We present simulation results showing that balanced push-pull mechanisms result in linear behavior, while imbalanced push-pull arrangements produce all of the incongruent DS-related behaviors that have been reported for simple cells.

Stimulus dependence of orientation and direction sensitivity of cat LGNd relay cells without cortical inputs: a comparison with area 17 cells

The cortical contribution to the orientation and direction sensitivity of LGNd relay cells was investigated by recording the responses of relay cells to drifting sinusoidal gratings of varying spatial frequencies, moving bars, and moving spots in cats in which the visual cortex (areas 17, 18, 19, and LS) was ablated. For comparison, the spatial-frequency dependence of orientation and direction tuning of striate cortical cells was investigated employing the same quantitative techniques used to test LGNd cells. There are no significant differences in the orientation and direction tuning to relay cells in the LGNd of normal and decorticate cats. The orientation and direction sensitivities of cortical cells are dependent on stimulus parameters in a fashion qualitatively similar to that of LGNd cells. The differences in the spatial-frequency bandwidths of LGNd cells and cortical cells may explain many of their differences in orientation and direction tuning. Although factors beyond narrowness of spatial-frequency tuning must exist to account for the much stronger orientation and direction preferences of cells in area 17 when compared to LGNd cells, the evidence suggests that the orientation and direction biases present in the afferents to the visual cortex may contribute to the orientation and direction selectivities found in cortical cells.