Neurons In Visual Areas Research Articles

Predictive coding networks for temporal prediction.

One of the key problems the brain faces is inferring the state of the world from a sequence of dynamically changing stimuli, and it is not yet clear how the sensory system achieves this task. A well-established computational framework for describing perceptual processes in the brain is provided by the theory of predictive coding. Although the original proposals of predictive coding have discussed temporal prediction, later work developing this theory mostly focused on static stimuli, and key questions on neural implementation and computational properties of temporal predictive coding networks remain open. Here, we address these questions and present a formulation of the temporal predictive coding model that can be naturally implemented in recurrent networks, in which activity dynamics rely only on local inputs to the neurons, and learning only utilises local Hebbian plasticity. Additionally, we show that temporal predictive coding networks can approximate the performance of the Kalman filter in predicting behaviour of linear systems, and behave as a variant of a Kalman filter which does not track its own subjective posterior variance. Importantly, temporal predictive coding networks can achieve similar accuracy as the Kalman filter without performing complex mathematical operations, but just employing simple computations that can be implemented by biological networks. Moreover, when trained with natural dynamic inputs, we found that temporal predictive coding can produce Gabor-like, motion-sensitive receptive fields resembling those observed in real neurons in visual areas. In addition, we demonstrate how the model can be effectively generalized to nonlinear systems. Overall, models presented in this paper show how biologically plausible circuits can predict future stimuli and may guide research on understanding specific neural circuits in brain areas involved in temporal prediction.

Spatial scaling of illusory motion perceived in a static figure.

In a phenomenon known as the Rotating Snakes illusion (Kitaoka & Ashida, 2003), illusory motion is perceived in a static figure with a specially designed luminance profile. It is known that the strength of this illusion increases with eccentricity, suggesting that the underlying mechanism of the illusion has a spatial property that changes with eccentricity. If a change in receptive-field size of responsible neurons causes the eccentricity dependence of the illusion, its strength should be spatially scalable using a scaling factor that increases with eccentricity, because the receptive field size of neurons in visual areas with retinotopy generally obeys quantitative dependence on eccentricity. For the luminance micropatterns comprising the figure for the Rotating Snakes illusion, we varied eccentricity from 9 to 15 deg and spatial frequency from 0.25 to 1.6 cycles/deg, and measured illusion strength. Illusion strength was found to increase with decreasing spatial frequency and with increasing eccentricity. Furthermore, the profiles of illusion strength at different eccentricities were spatially scalable into a single parabola as a function of the spatially scaled visual angle. The estimated scaling factors linearly increased with eccentricity with a slope similar to the eccentricity dependence of the receptive field size of V1 neurons, suggesting the involvement of early visual areas in the generation of the illusion.

Characterizing eye movement behaviors and kinematics of non-human primates during virtual navigation tasks.

Virtual environments (VE) allow testing complex behaviors in naturalistic settings by combining highly controlled visual stimuli with spatial navigation and other cognitive tasks. They also allow for the recording of eye movements using high-precision eye tracking techniques, which is important in electrophysiological studies examining the response properties of neurons in visual areas of nonhuman primates. However, during virtual navigation, the pattern of retinal stimulation can be highly dynamic which may influence eye movements. Here we examine whether and how eye movement patterns change as a function of dynamic visual stimulation during virtual navigation tasks, relative to standard oculomotor tasks. We trained two rhesus macaques to use a joystick to navigate in a VE to complete two tasks. To contrast VE behavior with classic measurements, the monkeys also performed a simple Cued Saccade task. We used a robust algorithm for rapid classification of saccades, fixations, and smooth pursuits. We then analyzed the kinematics of saccades during all tasks, and specifically during different phases of the VE tasks. We found that fixation to smooth pursuit ratios were smaller in VE tasks (4:5) compared to the Cued Saccade task (7:1), reflecting a more intensive use of smooth pursuit to foveate targets in VE than in a standard visually guided saccade task or during spontaneous fixations. Saccades made to rewarded targets (exploitation) tended to have increased peak velocities compared to saccades made to unrewarded objects (exploration). VE exploitation saccades were 6% slower than saccades to discrete targets in the Cued Saccade task. Virtual environments represent a technological advance in experimental design for nonhuman primates. Here we provide a framework to study the ways that eye movements change between and within static and dynamic displays.

Reward modulates the effect of visual cortical microstimulation on perceptual decisions.

Effective perceptual decisions rely upon combining sensory information with knowledge of the rewards available for different choices. However, it is not known where reward signals interact with the multiple stages of the perceptual decision-making pathway and by what mechanisms this may occur. We combined electrical microstimulation of functionally specific groups of neurons in visual area V5/MT with performance-contingent reward manipulation, while monkeys performed a visual discrimination task. Microstimulation was less effective in shifting perceptual choices towards the stimulus preferences of the stimulated neurons when available reward was larger. Psychophysical control experiments showed this result was not explained by a selective change in response strategy on microstimulated trials. A bounded accumulation decision model, applied to analyse behavioural performance, revealed that the interaction of expected reward with microstimulation can be explained if expected reward modulates a sensory representation stage of perceptual decision-making, in addition to the better-known effects at the integration stage.

Oblique effect in visual area 2 of macaque monkeys

The neural basis of an oblique effect, a reduced visual sensitivity for obliquely oriented stimuli, has been a matter of considerable debate. We have analyzed the orientation tuning of a relatively large number of neurons in the primary visual cortex (V1) and visual area 2 (V2) of anesthetized and paralyzed macaque monkeys. Neurons in V2 but not V1 of macaque monkeys showed clear oblique effects. This orientation anisotropy in V2 was more robust for those neurons that preferred higher spatial frequencies. We also determined whether V1 and V2 neurons exhibit a similar orientation anisotropy soon after birth. The oblique effect was absent in V1 of 4- and 8-week-old infant monkeys, but their V2 neurons showed a significant oblique effect. This orientation anisotropy in infant V2 was milder than that in adults. The results suggest that the oblique effect emerges in V2 based on the pattern of the connections that are established before birth and enhanced by the prolonged experience-dependent modifications of the neural circuitry in V2.

Attention is more than meets the eye

Our brains focus on important events and filter out distracting ones. An investigation in monkeys reveals a surprising dissociation between the neuronal and behavioural manifestations of attention. See Letter p.434 When we pay attention to certain locations, the responses of neurons in visual areas that represent those locations are enhanced. Here, Alexandre Zenon and Richard Krauzlis show that in adult rhesus monkeys, inactivation of the superior colliculus — part of the midbrain associated with eye movements — affects the perceptual effects of attention, but does not seem to affect the attentional modulation of upstream visual areas. Most work on attention is based on the idea that perceptual enhancement is driven by modulation of the visual response. This work suggests that visual attention involves other aspects of neuronal activity in the visual cortical areas — or perhaps other brain areas not directly associated with vision.

Local sensitivity to stimulus orientation and spatial frequency within the receptive fields of neurons in visual area 2 of macaque monkeys

We used dynamic dense noise stimuli and local spectral reverse correlation methods to reveal the local sensitivities of neurons in visual area 2 (V2) of macaque monkeys to orientation and spatial frequency within their receptive fields. This minimized the potentially confounding assumptions that are inherent in stimulus selections. The majority of neurons exhibited a relatively high degree of homogeneity for the preferred orientations and spatial frequencies in the spatial matrix of facilitatory subfields. However, about 20% of all neurons showed maximum orientation differences between neighboring subfields that were greater than 25 deg. The neurons preferring horizontal or vertical orientations showed less inhomogeneity in space than the neurons preferring oblique orientations. Over 50% of all units also exhibited suppressive profiles, and those were more heterogeneous than facilitatory profiles. The preferred orientation and spatial frequency of suppressive profiles differed substantially from those of facilitatory profiles, and the neurons with suppressive subfields had greater orientation selectivity than those without suppressive subfields. The peak suppression occurred with longer delays than the peak facilitation. These results suggest that the receptive field profiles of the majority of V2 neurons reflect the orderly convergence of V1 inputs over space, but that a subset of V2 neurons exhibit more complex response profiles having both suppressive and facilitatory subfields. These V2 neurons with heterogeneous subfield profiles could play an important role in the initial processing of complex stimulus features.

Effects of Brief Daily Periods of Unrestricted Vision during Early Monocular Form Deprivation on Development of Visual Area 2

Providing brief daily periods of unrestricted vision during early monocular form deprivation reduces the depth of amblyopia. To gain insights into the neural basis of the beneficial effects of this treatment, the binocular and monocular response properties of neurons were quantitatively analyzed in visual area 2 (V2) of form-deprived macaque monkeys. Beginning at 3 weeks of age, infant monkeys were deprived of clear vision in one eye for 12 hours every day until 21 weeks of age. They received daily periods of unrestricted vision for 0, 1, 2, or 4 hours during the form-deprivation period. After behavioral testing to measure the depth of the resulting amblyopia, microelectrode-recording experiments were conducted in V2. The ocular dominance imbalance away from the affected eye was reduced in the experimental monkeys and was generally proportional to the reduction in the depth of amblyopia in individual monkeys. There were no interocular differences in the spatial properties of V2 neurons in any subject group. However, the binocular disparity sensitivity of V2 neurons was significantly higher and binocular suppression was lower in monkeys that had unrestricted vision. The decrease in ocular dominance imbalance in V2 was the neuronal change most closely associated with the observed reduction in the depth of amblyopia. The results suggest that the degree to which extrastriate neurons can maintain functional connections with the deprived eye (i.e., reducing undersampling for the affected eye) is the most significant factor associated with the beneficial effects of brief periods of unrestricted vision.

Spatial non-homogeneity of the antagonistic surround in motion perception

At high contrast, duration thresholds for motion direction discrimination deteriorate with increasing stimulus size. This counterintuitive result has been explained by the center-surround antagonism present in the neurons of visual area MT. Conversely, at very low contrast, direction discrimination improves with increasing size, a result that has been explained by spatial summation. In this investigation, we study the effects of stimulus shape and contrast on center-surround antagonism. Using adaptive Bayesian staircases, we measured duration thresholds of 5 subjects for vertically oriented Gabor patches of 1 cycle/deg with two types of oval Gaussian windows, one vertically elongated (Sx = 0.35, Sy = 2.5 deg) and other horizontally elongated (Sx = 2.5, Sy = 0.35 deg) moving rightward or leftward at a speed of 2 deg/s. We found that at high contrast (92%) duration thresholds were lower for vertically than horizontally elongated windows. However, at low contrast (2.8%), we found that duration thresholds were lower for horizontally than vertically elongated windows. These asymmetric results mirror the spatial non-homogeneity of the antagonistic surround found in MT neurons and suggest that the underlying center-surround antagonism is stronger along the direction of motion.

Six principles of visual cortical dynamics.

A fundamental goal in vision science is to determine how many neurons in how many areas are required to compute a coherent interpretation of the visual scene. Here I propose six principles of cortical dynamics of visual processing in the first 150 ms following the appearance of a visual stimulus. Fast synaptic communication between neurons depends on the driving neurons and the biophysical history and driving forces of the target neurons. Under these constraints, the retina communicates changes in the field of view driving large populations of neurons in visual areas into a dynamic sequence of feed-forward communication and integration of the inward current of the change signal into the dendrites of higher order area neurons (30–70 ms). Simultaneously an even larger number of neurons within each area receiving feed-forward input are pre-excited to sub-threshold levels. The higher order area neurons communicate the results of their computations as feedback adding inward current to the excited and pre-excited neurons in lower areas. This feedback reconciles computational differences between higher and lower areas (75–120 ms). This brings the lower area neurons into a new dynamic regime characterized by reduced driving forces and sparse firing reflecting the visual areas interpretation of the current scene (140 ms). The population membrane potentials and net-inward/outward currents and firing are well behaved at the mesoscopic scale, such that the decoding in retinotopic cortical space shows the visual areas’ interpretation of the current scene. These dynamics have plausible biophysical explanations. The principles are theoretical, predictive, supported by recent experiments and easily lend themselves to experimental tests or computational modeling.

Beyond Poisson: Increased Spike-Time Regularity across Primate Parietal Cortex

Cortical areas differ in their patterns of connectivity, cellular composition, and functional architecture. Spike trains, on the other hand, are commonly assumed to follow similarly irregular dynamics across neocortex. We examined spike-time statistics in four parietal areas using a method that accounts for nonstationarities in firing rate. We found that, whereas neurons in visual areas fire irregularly, many cells in association and motor-like parietal regions show increasingly regular spike trains by comparison. Regularity was evident both in the shape of interspike interval distributions and in spike-count variability across trials. Thus, Poisson-like randomness is not a universal feature of neocortex. Rather, many parietal cells have reduced trial-to-trial variability in spike counts that could provide for more reliable firing-rate signals. These results suggest that spiking dynamics may play different roles in different cortical areas and should not be assumed to arise from fundamentally irreducible noise sources.

Orientation-specific adaptation to mentally generated lines in human visual cortex

Previous studies have shown that prolonged inspection of a tilted visual pattern leads to changes in perception (“tilt after-effect”, TAE), as well as to a reduction of the neural activation evoked by this pattern (“neural adaptation”). In this fMRI study, we investigated whether such perceptual and neural adaptation can be induced solely by mental imagination. The subjects were asked to mentally generate tilted lines, after which they were presented test lines oriented in the same or the direction orthogonal to the mentally tilted lines. Subjects showed a TAE even after mental imagery (“mental TAE”). Furthermore, a significant orientation-specific adaptation occurred in extrastriate visual areas (V3–V4), showing a decreasing gradient of adaptation from areas V4 to V1. Both this neural adaptation and the individual size of the mental TAE correlated with performance in a behavioural task probing mental imagery. Thus, orientation-selective neurons in visual areas seem to be recruited by mental imagery, and the amount of recruitment correlates with the degree of success of mental imagery operations. The influence of mental imagery on perception and on the neural activity in extrastriate visual areas provides evidence for analogue coding of mental images.

Postnatal Development of Disparity Sensitivity in Visual Area 2 (V2) of Macaque Monkeys

Macaque monkeys do not reliably discriminate binocular depth cues until about 8 wk of age. The neural factors that limit the development of fine depth perception in primates are not known. In adults, binocular depth perception critically depends on detection of relative binocular disparities and the earliest site in the primate visual brain where a substantial proportion of neurons are capable of discriminating relative disparity is visual area 2 (V2). We examined the disparity sensitivity of V2 neurons during the first 8 wk of life in infant monkeys and compared the responses of V2 neurons to those of V1 neurons. We found that the magnitude of response modulation in V2 and V1 neurons as a function of interocular spatial phase disparity was adult-like as early as 2 wk of age. However, the optimal spatial frequency and binocular response rate of these disparity sensitive neurons were more than an octave lower in 2- and 4-wk-old infants than in adults. Consequently, despite the lower variability of neuronal firing in V2 and V1 neurons of infant monkeys, the ability of these neurons to discriminate fine disparity differences was significantly reduced compared with adults. This reduction in disparity sensitivity of V2 and V1 neurons is likely to limit binocular depth perception during the first several weeks of a monkey's life.

The Effects of Visual Stimulation and Selective Visual Attention on Rhythmic Neuronal Synchronization in Macaque Area V4

Selective attention lends relevant sensory input priority access to higher-level brain areas and ultimately to behavior. Recent studies have suggested that those neurons in visual areas that are activated by an attended stimulus engage in enhanced gamma-band (30-70 Hz) synchronization compared with neurons activated by a distracter. Such precise synchronization could enhance the postsynaptic impact of cells carrying behaviorally relevant information. Previous studies have used the local field potential (LFP) power spectrum or spike-LFP coherence (SFC) to indirectly estimate spike synchronization. Here, we directly demonstrate zero-phase gamma-band coherence among spike trains of V4 neurons. This synchronization was particularly evident during visual stimulation and enhanced by selective attention, thus confirming the pattern inferred from LFP power and SFC. We therefore investigated the time course of LFP gamma-band power and found rapid dynamics consistent with interactions of top-down spatial and feature attention with bottom-up saliency. In addition to the modulation of synchronization during visual stimulation, selective attention significantly changed the prestimulus pattern of synchronization. Attention inside the receptive field of the recorded neuronal population enhanced gamma-band synchronization and strongly reduced alpha-band (9-11 Hz) synchronization in the prestimulus period. These results lend further support for a functional role of rhythmic neuronal synchronization in attentional stimulus selection.

Development of Temporal Response Properties and Contrast Sensitivity of V1 and V2 Neurons in Macaque Monkeys

The temporal contrast sensitivity of human infants is reduced compared to that of adults. It is not known which neural structures of our visual brain sets limits on the early maturation of temporal vision. In this study we investigated how individual neurons in the primary visual cortex (V1) and visual area 2 (V2) of infant monkeys respond to temporal modulation of spatially optimized grating stimuli and a range of stimulus contrasts. As early as 2 wk of age, V1 and V2 neurons exhibited band-pass temporal frequency tuning. However, the optimal temporal frequency and temporal resolution of V1 neurons were much lower in 2- and 4-wk-old infants than in 8-wk-old infants or adults. V2 neurons of 8-wk-old monkeys had significantly lower optimal temporal frequencies and resolutions than those of adults. Onset latency was longer in V1 at 2 and 4 wk of age and was slower in V2 even at 8 wk of age than in adults. Contrast threshold of V1 and V2 neurons was substantially higher in 2- and 4-wk-old infants but became adultlike by 8 wk of age. For the first 4 wk of life, responses to high-contrast stimuli saturated more readily in V2. The present results suggest that although the early development of temporal vision and contrast sensitivity may largely depend on the functional maturation of precortical structures, it is also likely to be limited by immaturities that are unique to V1 and V2.

Directional Bias of Neurons in V1 and V2 of Strabismic Monkeys: Temporal-to-Nasal Asymmetry?

Strabismus that develops shortly after birth is known to cause temporal-to-nasal eye movement asymmetries under monocular viewing. The neural mechanisms underlying this deficit are not well understood. In the current study, the hypothesis that this eye movement anomaly reflects a similar asymmetry in the directional response properties of neurons in the early stages of cortical processing was examined. Strabismus was simulated with optical methods in infant monkeys between 4 and 14 weeks of age. When the monkeys were mature, microelectrode recording experiments were conducted in the primary visual cortex (V1) and visual area 2 (V2). After the spatial frequency of sine wave-grating stimuli for each neuron was optimized, each neuron's responsiveness to 24 directions of stimulus movement was measured. The preferred direction and the strength of directional bias were determined by a vector summation method: There was not an overabundance of neurons in V1 or V2 of strabismic monkeys preferring the temporal-to-nasal direction of stimulus movement. However, the average directional bias was significantly reduced in these strabismic monkeys. Interocular suppression was highly prevalent, and this suppression was stronger and more common in neurons dominated by the ipsilateral eye. The results suggest that the eye movement asymmetries in strabismic subjects do not result from similar asymmetries in the directional properties of V1 or V2 neurons, but rather reflect impoverished cortical signals to the brain stem nuclei that control eye movements.

Delayed maturation of receptive field center/surround mechanisms in V2

Neurons in the adult visual cortex are capable of integrating signals over a large area that surrounds their classic receptive field (RF), and this ability of cortical neurons is thought to be intimately involved in perceptual binding. It is not known, however, at what age these long-range signal interactions emerge. Here, we report that qualitatively adult-like center/surround interactions are already present in the primary visual cortex as early as postnatal day 14 in macaque monkeys. However, the RF surrounds of visual area 2 (V2) neurons were largely absent until 4 weeks of age and, as late as 8 weeks of age, center/surround signal interactions in V2 neurons were immature. Our results suggest that the cortical circuits underlying the RF center/surround of individual neurons mature considerably later in V2 than in the primary visual cortex and give critical evidence for the hypothesis that the functional maturation of the primate visual brain proceeds in a hierarchical manner.

Functional organization of speed tuned neurons in visual area MT.

We analyzed the functional organization of speed tuned neurons in extrastriate visual area MT. We sought to determine whether neurons tuned for particular speeds are clustered spatially and whether such spatial clusters are elongated normal to the cortical surface so as to form speed columns. Our data showed that MT neurons are indeed clustered according to preferred speed. Multiunit recordings were speed tuned, and the speed tuning of these signals was well correlated with the speed tuning of single neurons recorded simultaneously. To determine whether speed columns exist in MT, we compared the rates at which preferred speed changed in electrode tracks that traversed MT obliquely and normally to the cortical surface. If speed columns exist, the preferred speed should change at a faster rate during oblique electrode tracks. We found, however, that preferred speed changed at similar rates for either type of penetration. In the same data set, the rate of change of preferred direction and preferred disparity differed substantially in normal and oblique penetrations as expected from the known columnar organization of MT. Thus our results suggest that a columnar organization for speed tuned neurons does not exist in MT.

INVARIANT FACE AND OBJECT RECOGNITION IN THE VISUAL SYSTEM

Neurophysiological evidence is described, showing that some neurons in the macaque temporal cortical visual areas have responses that are invariant with respect to the position, size and view of faces and objects, and that these neurons show rapid processing and rapid learning. A theory is then described of how such invariant representations may be produced in a hierarchically organized set of visual cortical areas with convergent connectivity. The theory proposes that neurons in these visual areas use a modified Hebb synaptic modification rule with a short-term memory trace to capture whatever can be captured at each stage that is invariant about objects as the object changes in retinal position, size, rotation and view. Simulations are then described which explore the operation of the architecture. The simulations show that such a processing system can build invariant representations of objects. © 1977 Elsevier Science Ltd. All Rights Reserved.

Basic information processing and higher cognition: does the mammalian cerebral cortex deal with them at different hierarchical levels?

1. The primary visual cortex has been generally considered as a feature processing area where the neurons code for basic stimulus attributes, such as orientation, colour and direction of movement. Higher cognitive functions, such as attention and memory, were believed to be mediated by extrastriate regions at higher levels in a hierarchy of visual areas. However, the experiments reported here suggest that the function of the primary visual cortex is more than basic information processing in at least two respects. 2. First, significant modulation of visual responses occur in the primary visual cortex (V1) of the monkey depending upon attentional factors. There appears to be a feedback to V1 from extrastriate regions that facilitates neuronal responses in areas of attention and suppresses responses in other regions of the visual field. 3. Second, one would expect a region that is concerned with coding basic stimulus features to be rather resistant to plastic changes, as cells need to be reliable feature detectors for meaningful behaviour by the organism. However, the experiments on anaesthetized cats reported here show that a subset of neurons in visual area 17 of the cat have a potential for synaptic plasticity. Their response properties can undergo long-term changes under certain circumstances. To induce these changes, specific visual stimuli were combined with iontophoretic administration of NMDA and the GABA antagonist, bicuculline, which are believed to provide the molecular milieu for long-term potential and long-term depression.