Search

CONSPICUOUS visual features commonly attract gaze1,2, but how the brain selects targets for eye movements is not known. We investigated target selection in rhesus monkeys performing a visual search task3 by recording neurons in the frontal eye field, an area known to be responsible for generating purposive eye movements4,5. Neurons with combined visual- and eye movement-related activity were analysed. We found that the initial visual responses to search stimulus arrays were the same whether the target or a distractor was in the response field. We also found that the neural activity evolved to specify target location before the execution of eye movements, ultimately peaking when the target was in the response field and being suppressed when the target was beside but not distant from the response field. These results demonstrate a possible mechanism by which a desired target is fixated and inappropriate eye movements are prevented.

Two discrete areas in frontal cortex are involved in generating saccadic eye movements—the frontal eye field (FEF) and the supplementary eye field (SEF). Whereas FEF represents saccades in a topographic retinotopic map, recent evidence indicates that saccades may be represented craniotopically in SEF. To further investigate the relationship between these areas, the topographic organization of afferents to FEF from SEF in Macaco mulatto was examined by placing injections of distinct retrograde tracers into different parts of FEF that represented saccades of different amplitudes. Central FEF (lateral area 8A), which represents saccades of intermediate amplitudes, received afferents from a larger portion of SEF than did lateral FEF (area 45), which represents shorter saccades, or medial FEF (medial area 8A), which represents the longest saccades in addition to pinna movements. Moreover, in every case the zone in SEF that innervated lateral FEF (area 45) also projected to medial FEF (area 8A). In one case, a zone in rostral SEF projected to both lateral area 8A from which eye movements were evoked by microstimulation as well as medial area 8A from which pinna movements were elicited by microstimulation. This pattern of afferent convergence and divergence from SEF onto the retinotopic saccade map in FEF is indicative of some sort of map transformation between SEF and FEF. Such a transformation would be necessary to interconnect a topographic craniotopic saccade representation in SEF with a topographic retinotopic saccade representation in FEF.

The degree of parallel processing in frontal cortex-basal ganglia circuits is a central and debated issue in research on the basal ganglia. To approach this issue directly, we analyzed and compared the corticostriatal projections of two principal oculomotor areas of the frontal lobes, the frontal eye field (FEF) and the supplementary eye field (SEF). We first identified cortical regions within or adjacent to each eye field by microstimulation in macaque monkeys and then injected each site with either 35S-methionine or WGA-HRP conjugate. We analyzed the corticostriatal projections and also the interconnections of the pairs of cortical areas. We observed major convergence of the projections of the FEF and the SEF within the striatum, principally in the caudate nucleus. In cross sections through the striatum, both projections were broken into a series of discontinuous input zones that seemed to be part of complex three-dimensional labyrinths. Where the FEF and SEF projection fields were both present, they overlapped patch for patch. Thus, both inputs were dispersed within the striatum but converged with one another. Striatal afferents from cortex adjacent to the FEF and the SEF did not show convergence with SEF and FEF inputs, but did, in part, converge with one another. For all pairs of cortical areas tested, the degree of overlap in the corticostriatal projections appeared to be directly correlated with the degree of cortical interconnectivity of the areas injected. All of the corticostriatal fiber projections observed primarily avoided immunohistochemically identified striosomes. We conclude that there is convergence of oculomotor information from two distinct regions of the frontal cortex to the striatal matrix, which is known to project into pallidonigral circuits including the striatonigrocollicular pathway of the saccadic eye movement system. Furthermore, functionally distinct premotor areas near the oculomotor fields often systematically projected to striatal zones adjacent to oculomotor field projections, suggesting an anatomical basis for potential interaction of these inputs within the striatum. We propose that parallel processing is not the exclusive principle of organization of forebrain circuits associated with the basal ganglia. Rather, patterns of both convergence and divergence are present and are likely to depend on multiple functional and developmental constraints.

1. The purpose of this study was to analyze the response properties of neurons in the frontal eye fields (FEF) of rhesus monkeys (Macaca mulatta) and to compare and contrast the various functional classes with those recorded in the supplementary eye fields (SEF) of the same animals performing the same go/no-go visual tracking task. Three hundred ten cells recorded in FEF provided the data for this investigation. 2. Visual cells in FEF responded to the stimuli that guided the eye movements. The visual cells in FEF responded with a slightly shorter latency and were more consistent and phasic in their activation than their counterparts in SEF. The receptive fields tended to emphasize the contralateral hemifield to the same extent as those observed in SEF visual cells. 3. Preparatory set cells began to discharge after the presentation of the target and ceased firing before the saccade, after the go/no-go cue was given. These neurons comprised a smaller proportion in FEF than in SEF. In contrast to their counterparts in SEF, the preparatory set cells in FEF did not respond preferentially in relation to contralateral movements, even though most responded preferentially for movements in one particular direction. The time course of the discharge of the FEF set cells was similar to that of their SEF counterparts, except that they reached their peak level of activation sooner. The few preparatory set cells in FEF tested with both auditory and visual stimuli tended to respond preferentially to the visual targets, whereas, in contrast, most set cells in SEF were bimodal. 4. Sensory-movement cells represented the largest population of cells recorded in FEF, responding in relation to both the presentation of the targets and the execution of the saccade. Although some of these sensory-movement cells resembled their counterparts in SEF by exhibiting a sustained elevation of activity, most of the FEF sensory-movement cells gave two discrete bursts, one after the presentation of the target and another before and during the saccade. Like their counterparts in SEF, the sensory-movement cells tended to be tuned for saccades into the contralateral hemifield, but this tendency was more pronounced in FEF than in SEF. The FEF sensory-movement cells discharged more briskly, with a shorter latency relative to the presentation of the target, than their counterparts in SEF. In addition, the FEF sensory-movement neurons reached their peak activation sooner than SEF sensory-movement neurons. Most FEF sensory-movement cells exhibited different patterns of activation in response to visual and auditory targets.(ABSTRACT TRUNCATED AT 400 WORDS)

Kim and Shadlen investigate how neurons of the prefrontal cortex interpret slowly accumulating signals from visual cortex to make a perceptual decision.

When the two eyes are exposed to markedly different patterns, perception becomes unstable, falling into oscillations, so that the image of one eye is seen first and then that from the other. With large stimuli the alternation is piecemeal, whilst when small stimuli are used the whole pattern alternates in unison. The purpose of this study was to determine whether a reliable, objective indicator of the perceptual state during binocular rivalry could be developed in the nonhuman primate. Monkeys (Macaca mulatta) were trained to discriminate direction of motion when presented with vertically drifting gratings moving in opposite directions in the two eyes. A high correlation was found between the direction of the slow phase of the optokinetic nystagmus (OKN) elicited by the drifting gratings during rivalry and the direction of motion reported by the monkey even though the gain of the OKN was reduced during rivalry, and the latency was longer. Behavioral eye dominance during rivalry varied significantly over time, between individuals and as a function of interocular contrast differences. Since the direction of tracking eye movements can be used to reliably monitor perceptual state during binocular motion rivalry, the opportunity exists in nonhuman primates to study the neurophysiological mechanisms underlying motion perception during the perceptually ambiguous condition of binocular rivalry.

Buschman and Miller (Reports, 30 March 2007, p. 1860) described the activity of ensembles of neurons in parietal and frontal cortex of monkeys performing visual search for targets that were easy or hard to distinguish from distractors. However, their conclusions are called into question by discrepancies between their results and publications from other laboratories measuring the same neural process.

Retinal ganglion cells were studied in pigmented ferrets that received small electrophoretic injections of horseradish peroxidase (HRP) into the dorsal lateral geniculate nucleus (LGNd) or optic tract. Ferret retina contains a number of types of retinal ganglion cells of which the relative cell body sizes, dendritic field structures, and central projections correspond closely to those of retinal ganglion cell types in the cat. Ferret retina contains about the same proportion of alphalike cells, a lower proportion of betalike cells, and thus a high proportion of other types of ganglion cells than cat retina. Ferret retina has a visual streak and somewhat weaker area centralis than cat retina. Changes in ganglion cell morphology associated with eccentricity are less pronounced in the ferret than in the cat. The adult ferret retina is about 12.5 mm in diameter, and the nasotemporal division is about 2.7 mm from the temporal margin. Interestingly, virtually all alpha cells in the pigmented ferrets studied projected contralaterally. Studies of infant ferrets indicate that 4 days after birth (P4) the area of ferret retina is 25% that of the adult. The neonatal ferret retina contains numerous small, densely packed cells in the presumptive ganglion cell layer. At P4 these cells appear to be uniformly distributed across the retina. The area centralis and visual streak are not obvious as late as 8 days after birth.

We analyzed the dendritic field morphology of 297 ganglion cells from peripheral regions of monkey retina. Most of the dendritic fields were elongated, and there was a significant tendency for the dendritic fields to be oriented radially, i.e., like the spokes of a wheel with the fovea at the hub. An overrepresentation of radial orientations in the peripheral retina of primates might explain why humans are best able to detect stimuli which are oriented radially using peripheral vision.

The morphology of ganglion cell dendritic trees varies across the cat retina. Evidence is presented that the variation in two attributes of ganglion cell dendritic structure can be accounted for by specific aspects of the topography of the adult and developing retina. The first attribute considered was the displacement of the center of the dendritic field from the cell body in the plane of the retina. The results of this study provide evidence that most ganglion cell dendritic fields are displaced away from neighboring cells, i.e., down the retinal ganglion cell density gradient. Because of the systematic dendritic displacement locally, the centers of the dendritic fields are arranged in a more precise mosaic than are their cell bodies. The second attribute considered was the elongation and orientation of the dendritic fields. From approximately embryonic day 50 to postnatal day 10 the cat retina undergoes a process of maturation (reviewed by Rapaport and Stone: Neuroscience 11:289–301, '84) that begins at the area centralis and spreads over the retina in a horizontally elongated wave. We found that the elongation and orientation of retinal ganglion cell dendritic fields is significantly correlated with the shape of the wave of maturation. The orientation of a dendritic field is not predicted by the direction of its displacement nor is it directly related to the distribution of neighboring retinal ganglion cells. These results indicate that the displacement of a ganglion cell's dendritic field from its cell body results from mechanisms different from those responsible for the orientation of the dendritic field. Factors that may be responsible for these two attributes of ganglion cell dendritic morphology are discussed.

The degree to which a retinal ganglion cell's environment can affect its morphological development was studied by manipulating the distribution of ganglion cells in the developing cat retina. In the newborn kitten there is an exuberant ganglion cell projection from temporal retina to the contralateral lateral geniculate nucleus (LGNd) (Leventhal et al., 1988) and from nasal retina to the ipsilateral LGNd. Neonatal, unilateral optic tract section results in the survival of many of these ganglion cells (Leventhal et al., 1988). The morphology of ganglion cells which survive in regions of massively reduced ganglion cell density was studied. As reported previously (Linden and Perry, 1982; Perry and Linden, 1982; Ault et al., 1985; Eysel et al., 1985), we found that the dendritic fields of all types of ganglion cells on the border of an area depleted of ganglion cells extended into the depleted area. The cell bodies and dendritic fields of alpha and beta cells within depopulated areas, as well as on the borders of the depopulated areas, were larger than normal. The dendritic fields of these cells also exhibited abnormal branching patterns. For alpha and beta cell types the relative increase in size tended to be greatest where the relative change in density was the greatest. In fact, isolated beta cells within the cell-poor area centralis region resembled normal central alpha cells in the cell-rich region of the area centralis in the same retina. Interestingly, in the same regions of reduced density where alpha and beta cells were dramatically larger than normal, the cell body and dendritic field sizes of other cell types (epsilon, g1 and g2 were unchanged. These results indicate that neuronal interactions during development contribute to the morphological differentiation of retinal ganglion cells and that different mechanisms mediate the morphological development of different classes of cells in cat retina.

The development of the nasotemporal division in cat retina was studied. We find that in the normally pigmented neonatal cat significant numbers of ganglion cells of all types in temporal retina project to the contralateral dorsal lateral geniculate nucleus (LGNd); far fewer cells in temporal retina project contralaterally to the LGNd in the normal adult. Thus, most of these cells must be eliminated during development. Experimental interruption of one optic tract in the neonate results in the retrograde degeneration of the ipsilaterally projecting ganglion cells in the temporal retina ipsilateral to the lesion. Consequent to the loss of the ipsilaterally projecting cells in this hemiretina, many of the ganglion cells projecting to the intact contralateral LGNd, which are normally eliminated, survive. Also, unlike in the normal cat, in which very few of the small ganglion cells in temporal retina project contralaterally to the thalamus, in optic tract sectioned (OTX) cats, significant numbers of the smallest ganglion cells in the temporal retina ipsilateral to the lesion project contralaterally to the intact thalamus. In order to make a quantitative comparison of the distributions of ipsilaterally and contralaterally projecting cells in the temporal retinae of normal cats, OTX cats, and neonatal kittens, it was necessary to determine the position of the vertical meridian in all animals. We defined the vertical meridian as the median edge (Stone, 1966). The median edge was determined from the distribution of the most nasally located, ipsilaterally projecting cells in temporal retina. The results indicate that the angle of the vertical meridian (median edge) with respect to the area centralis and optic disc is specified before birth and does not differ in normal cats, OTX cats, or neonatal kittens. Since the location of the vertical meridian does not change with age in postnatal life and is not affected by optic tract section, corresponding regions of retina in the different groups could be compared. A quantitative analysis of ganglion cell density in the temporal retina contralateral to the section, ipsilateral to the intact hemisphere, indicated that there was a reduction in the population of ipsilaterally projecting ganglion cells that was complementary to the abnormally large number of contralaterally projecting cells surviving in the temporal retina ipsilateral to the lesion.(ABSTRACT TRUNCATED AT 400 WORDS)

The dendritic field size, the distribution of the dendrites relative to the cell body, and the overall shape of the dendritic field of type I ganglion cells in the rat retina were analyzed. These features of neuronal structure were related to the topography of the rat retina. As in the cat, the cell bodies of type I ganglion cells are arranged in a nonrandom mosaic. Previous work has demonstrated that the density of type I cells in the rat retina does not covary with the density of all ganglion cells. Type I dendritic field size varies over the retina; the increase in dendritic field size is accounted for better by the decrease in type I density than by the decrease in overall ganglion cell density. The center of the dendritic field of most type I cells is displaced in the plane of the retina from the cell body. Unlike in carnivore retina (Schall and Leventhal: J. Comp. Neurol. 257:149-159, '87), the dendritic fields in the rat are not displaced down the ganglion cell density gradient. Rather, there is a tendency for the dendritic trees, especially in temporal retina, to be displaced toward dorsal retina. Most of the dendritic fields are elongated, but the degree of elongation is less than that observed in carnivore or primate retina. Unlike in carnivore and primate retina (Leventhal and Schall: J. Comp. Neurol. 220:465-475, '83; Schall et al.: Brain Res. 368:18-23, '86), there is no relationship between dendritic tree orientation and position relative to any point on the retina in the rat.(ABSTRACT TRUNCATED AT 250 WORDS)

Most retinal ganglion cells (Levick and Thibos, 1982) and cortical cells (Leventhal, 1983; Leventhal et al., 1984) subserving peripheral vision respond best to stimuli that are oriented radially, i.e., like the spokes of a wheel with the area centralis at the hub. We have extended this work by comparing directly the distributions of orientations represented in topographically corresponding regions of retina and visual cortex. Both central and peripheral regions were studied. The relations between the orientations of neighboring ganglion cells and the manner in which the overrepresentation of radial orientations is accommodated in the functional architecture of visual cortex were also studied. Our results are based on an analysis of the orientations of the dendritic fields of 1296 ganglion cells throughout the retina and the preferred orientations of 1389 cells located in retinotopically corresponding regions of cortical areas 17, 18, and 19 in the cat. We find that horizontal and vertical orientations are overrepresented in regions of both retina and visual cortex subserving the central 5 degrees of vision. The distributions of the orientations of retinal ganglion cells and cortical cells subserving the horizontal, vertical, and diagonal meridians outside the area centralis differ significantly. The distribution of the preferred orientations of the S (simple) cells in areas 17, 18 and 19 subserving a given part of the retina corresponds to the distribution of the dendritic field orientations of the ganglion cells in that part of retina. The distribution of the preferred orientations of C (complex) cells with narrow receptive fields in area 17 but not C cells with wide receptive fields in areas 17, 18, or 19 subserving a given part of the retina matches the distribution of the orientations of the ganglion cells in that part of retina. The orientations of all of the alpha-cells in 5–9 mm2 patches of retina along the horizontal, vertical, and oblique meridians were determined. A comparison of the orientations of neighboring cells indicates that other than a mutual tendency to be oriented radially, ganglion cells with similar orientations are not clustered in the retina. Reconstructions of electrode penetrations into regions of visual cortex representing peripheral retina indicate that columns subserving radial orientations are wider than those subserving nonradial orientations. Our results provide evidence that the distribution of the preferred orientations of simple cells in visual cortex subserving any region of the visual field matches the distribution of the orientations of the ganglion cells subserving the same region of the visual field.(ABSTRACT TRUNCATED AT 400 WORDS)

Two manipulations of a visual search task were used to test the hypothesis that the discrimination of a target from distractors by visually responsive neurons in the frontal eye field (FEF) marks the outcome and conclusion of visual processing instead of saccade preparation. First, search efficiency was reduced by increasing the similarity of the distractors to the target. Second, response interference was introduced by infrequently changing the location of the target in the array. Both manipulations increased reaction time, but only the change in search efficiency affected the time needed to select the target by visually responsive neurons. This result indicates that visually responsive neurons in FEF form an explicit representation of the location of the target in the image.

We investigated how saccade target selection by humans and macaque monkeys reacts to unexpected changes of the image. This was explored using double step and search step tasks in which a target, presented alone or as a singleton in a visual search array, steps to a different location on infrequent, random trials. We report that human and macaque monkey performance are qualitatively indistinguishable. Performance is stochastic with the probability of producing a compensated saccade to the final target location decreasing with the delay of the step. Compensated saccades to the final target location are produced with latencies relative to the step that are comparable to or less than the average latency of saccades on trials with no target step. Noncompensated errors to the initial target location are produced with latencies less than the average latency of saccades on trials with no target step. Noncompensated saccades to the initial target location are followed by corrective saccades to the final target location following an intersaccade interval that decreases with the interval between the target step and the initiation of the noncompensated saccade. We show that this pattern of results cannot be accounted for by a race between two stochastically independent processes producing the saccade to the initial target location and another process producing the saccade to the final target location. However, performance can be accounted for by a race between three stochastically independent processes—a GO process producing the saccade to the initial target location, a STOP process interrupting that GO process, and another GO process producing the saccade to the final target location. Furthermore, if the STOP process and second GO process start at the same time, then the model can account for the incidence and latency of mid-flight corrections and rapid corrective saccades. This model provides a computational account of saccade production when the image changes unexpectedly. � 2007 Elsevier Ltd. All rights reserved.

In the normally pigmented neonatal cat, many ganglion cells in temporal retina project to the contralateral dorsal lateral geniculate nucleus (LGNd) and medial interlaminar nucleus (MIN). Most of these cells are eliminated during postnatal development. If one optic tract is sectioned at birth, much of this exuberant projection from the contralateral temporal retina is stabilized (Leventhal et al., 1988b). To determine how the abnormal projection from the contralateral temporal retina is accommodated in the central visual pathways, neuronal activity was recorded in the visual thalamus and cortex of adult cats whose optic tracts were sectioned as neonates. The recordings showed that up to 20 degrees of the ipsilateral hemifield is represented in the LGNd and MIN. Recordings from areas 17 and 18 of the intact visual cortex showed that up to 20 degrees of the ipsilateral visual field is also represented and that the ipsilateral representation is organized as in a Boston Siamese cat (Hubel and Wiesel, 1971; Shatz, 1977; Cooper and Blasdel, 1980) or a heterozygous albino cat (Leventhal et al., 1985b). The extent of the ipsilateral visual field representation was greater in area 18 than in area 17; the extent of the ipsilateral hemifield representation in areas 17 and 18 varied with elevation, increasing with distance from the horizontal meridian. The receptive fields of cells in the LGNd and visual cortex subserving contralateral temporal retina were abnormally large. Otherwise, their receptive field properties seemed normal. In the same animals studied physiologically, HRP was injected into the ipsilateral hemifield representation in the LGNd and MIN of the intact hemisphere. The topographic distribution of the alpha and beta cells, respectively, labeled by these injections correlated with the elevation-related changes in the ipsilateral visual field representation in areas 18 and 17. Our results indicate that the retinotopic organization of the mature geniculocortical pathway reflects the abnormal pattern of central projections of ganglion cells in neonatally optic tract sectioned cats. Thus, if they do not die, retinal ganglion cells normally eliminated during development are capable of making seemingly normal, functional connections. The finding that an albino-like representation of the ipsilateral hemifield can be induced in the visual cortex of normally pigmented cats suggests that the well- documented defects in the geniculocortical pathways of albinos are secondary to the initial misrouting of ganglion cells at the optic chiasm (Kliot and Shatz, 1985) and not a result of albinism per se.

Neuronal activity in the superior temporal sulcus of monkeys, a cortical region that plays an important role in analyzing visual motion, was related to the subjective perception of movement during a visual task. Single neurons were recorded while monkeys (Macaca mulatta) discriminated the direction of motion of stimuli that could be seen moving in either of two directions during binocular rivalry. The activity of many neurons was dictated by the retinal stimulus. Other neurons, however, reflected the monkeys' reported perception of motion direction, indicating that these neurons in the superior temporal sulcus may mediate the perceptual experience of a moving object.

Orientation sensitivity is a characteristic of most retinal ganglion cells (Levick and Thibos, '82), most relay cells in the dorsal lateral geniculate nucleus (Vidyasagar and Urbas, '82), and most neurons in the visual cortex (Hubel and Wiesel, '62) in the cat. In the retina there is a systematic relationship between receptive field position (polar angle) and preferred orientation. Outside of the area centralis most retinal ganglion cells respond best to stimuli oriented radially, i.e., oriented parallel to the line connecting their receptive fields to the area centralis (Levick and Thibos, '82). This relationship is strongest along the horizontal meridian (the visual streak) and appears to reflect the innate, radial orientation of retinal ganglion cell dendritic fields (Leventhal and Schall, '83). A relationship between preferred orientation and polar angle also exists in cat striate cortex; outside of the area centralis representation most cells respond best to lines oriented radially. This relationship is strongest for S type cells, the most orientation-selective cells, and cells in regions representing the horizontal meridian (Leventhal, '83). To determine if similar relationships exist in cat extrastriate cortex, the preferred orientations and receptive field positions of 226 neurons in area 19 were studied. We find that, as in area 17, most area 19 cells outside of the representation of the area centralis respond best to lines oriented radially; this relationship is strongest for the cells having the narrowest receptive fields and in regions subserving the horizontal meridian. Unlike in striate cortex, in area 19 the relationship between preferred orientation and polar angle is not dependent upon cell type (S or C) or to the degree of orientation sensitivity exhibited. Also, in area 19, but not in area 17, the relationship between preferred orientation and polar angle fails for the cells having the widest receptive fields. The results of this study show that the systematic relationship between preferred orientation and receptive field position which begins in the retina is largely preserved through extrastriate cortex. The functional architecture of orientation sensitive cells throughout the visual pathways may depend ultimately upon the innate, radial orientation of the dendritic fields of retinal ganglion cells.

We investigated the structural basis of the physiological orientation sensitivity of retinal ganglion cells (Levick and Thibos, '82). The dendritic fields of 840 retinal ganglion cells labeled by injections of horseradish peroxidase into the dorsal lateral geniculate nucleus (LGNd) or optic tracts of normal cats, Siamese cats, and cat deprived of patterned visual experience from birth by monocular lid-suture (MD) were studied. Mathematical techniques designed to analyze direction were used to find the dendritic field orientation of each cell. Statistical techniques designed for angular data were used to determine the relationship between dendritic field orientation and angular position on the retina (polar angle). Our results indicate that 88% of retinal ganglion cells have oriented dendritic fields and that dendritic field orientation is related systematically to retinal position. In all regions of retina more than 0.5 mm from the area centralis the dendritic fields of retinal ganglion cells are oriented radially, i.e., like the spokes of a wheel having the area centralis at its hub. This relationship was present in all animals and cell types studied and was strongest for cells located close to the horizontal meridian (visual streak) of the retina. Retinal ganglion cells appear to be sensitive to stimulus orientation because they have oriented dendritic fields.

Recent studies have measured the time taken by neural processes to decide between alternative stimuli. Overt responses are produced only as soon as neural processes are completed. The brain throttles the mind.