| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Department of Neurobiology and Anatomy and Center for Visual Science, University of Rochester, Rochester, New York 14642
Submitted 5 September 2003; accepted in final form 27 September 2003
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONDISCLOSURESACKNOWLEDGMENTSREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONDISCLOSURESACKNOWLEDGMENTSREFERENCES |
|---|
; Fuster and Alexander 1971; Miller and Desimone 1994). However, the exact roles of intermediate visual cortical areas remain unclear, despite mounting evidence that visual processing and working memory are mediated by a distributed network of neurons (for review see Fuster 1995b; Goldman-Rakic 1995; Miller and Cohen 2001).
The middle temporal (MT) area is a midlevel region in primate cortex thought to be a major site of visual motion processing in the primate brain (for recent review see Pasternak et al. 2003). Its neurons have been shown to play a role in a number of behavioral tasks involving motion discrimination (Bisley et al. 2001; Britten et al. 1992, 1996; Ditterich et al. 2003; Nichols and Newsome 2002; Salzman et al. 1992) and their loss results in deficits in motion perception (for recent review see Merigan and Pasternak 2002). Recently, a series of studies used psychophysical (Pasternak and Zaksas 2003; Zaksas et al. 2001), microstimulation (Bisley et al. 2001), and lesion (Bisley and Pasternak 2000) approaches to examine the role of MT in the performance of working memory tasks involving visual motion. These studies provided evidence that MT neurons may be involved not only in processing of visual motion but also in its temporary storage.
The present study examined the activity of MT neurons during a match-to-sample task. We were particularly interested in neuronal activity during the memory period and whether this activity reflects the properties of the remembered stimulus. We recorded from MT neurons while monkeys compared the directions of motion of 2 stimuli, sample and test, separated by a brief delay. We found that the activity of many MT neurons did not return to baseline during the delay. Rather, they showed a characteristic pattern consisting of a small burst of activity early in the delay, followed by a period of suppression and subsequent increase in firing rate. The activity throughout most of the delay, but especially in the early delay, was directionally biased in a way that could not be entirely explained by adaptation. Together with previous studies, these data suggest that MT neurons both process the information about visual motion and may be involved in the circuitry supporting temporary storage of this information.
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONDISCLOSURESACKNOWLEDGMENTSREFERENCES |
|---|
Recordings were performed in 2 adult macaque monkeys (Macaca nemestrina) weighing about 8 and 9 kg. On weekdays, water was restricted for a period of 22 h before testing and the daily water ration, in the form of a fruit drink, was provided during the behavioral testing. On weekends, the monkeys were not tested behaviorally and received 100 ml/kg water per day. Food was continually available in the home cage and monkeys received supplements of fresh fruit and vitamins daily. Body weights were recorded at least 3 times/wk to ensure good health and normal growth. The monkeys were implanted with scleral search coils and head-restraint devices to monitor their eye position, and had recording cylinders placed above the superior temporal sulcus (STS). Before the present study, these monkeys were tested on a variety of visual discrimination tasks involving random-dot stimuli. The results of these measurements have been published elsewhere (Bisley et al. 2001; Pasternak and Zaksas 2003; Zaksas et al. 2001). Experiments were carried out in accordance with the guidelines published in the National Institutes of Health Guide for the Care and Use of Laboratory Animals (National Institutes of Health publication no. 86-23, revised 1987) and were approved by the University of Rochester Committee for Animal Research.
Stimuli
Visual stimuli were presented on a video display (76-Hz frame rate) located 42 cm in front of the monkey. The random-dot stimuli were identical to those used in previous studies (Bisley and Pasternak 2000; Rudolph and Pasternak 1999) and consisted of dots placed randomly within a circular aperture, the size of which was matched to the receptive field size of the recorded neuron. Each dot was displaced by a constant step size (
x) and temporal interval (t = 13 ms). The lifetime of an individual dot was equal to the duration of the stimulus presentation (500 ms) and in each frame the dots moved independently in any direction chosen at random from a uniform distribution. This distribution could range from 0° (all the dots moving in parallel) to 360° (only random motion; see Fig. 1A). This value is termed the direction range of the stimulus. The dots, viewed at a distance of 42 cm, were 0.03° in diameter, and their luminance was set about 3.5 log units above human detection threshold. Optimal dot density and speed were determined for a given neuron and those values were used during behavioral testing. The dot densities ranged from 2 to 5 dots/deg2 and the velocities varied from 3°/s (x = 0.03) to 35°/s (x = 0.45).
|
We used a version of a working memory task, first introduced by Konorski (1959), in which the monkeys compared the directions of 2 moving stimuli separated by a delay and reported them as the same or different. The monkeys initiated each trial by fixating a small spot for 1,000 ms within a 1.5° window. Two random-dot stimuli, the sample and test, then appeared sequentially, separated by a 1.5- or 3-s delay (Fig. 1B). Both the sample and test stimuli were presented within the receptive field for 500 ms. After the termination of the test, the monkeys had 3 s to respond by pressing one of 2 buttons—signaling either that the net direction of motion was the same or opposite in the 2 stimuli. On each trial, the direction range of the sample was selected at random from 4 values (usually 0, 150, 300, and 360°), keeping performance at approximately 80% correct. The test stimulus always consisted of coherently moving dots (0° range) that moved in either the same or the opposite direction. Because most of the analysis was performed on trials containing a coherently moving sample stimulus (0° range) the number of such trials was maximized by presenting them twice as often as the other 3 range values. On trials with the 360°-range sample (i.e., that with no net direction of motion) the animals were rewarded at random.
Surgical procedures
After the monkeys had learned the basic behavioral task, they received scleral search coils and head-restraint devices for monitoring eye position (Remmel 1984). For all surgical procedures, which were performed under aseptic conditions, the animals were initially anesthetized with ketamine hydrochloride (15 mg/kg intramuscularly) and maintained with 3% isoflurane. Craniotomies were made over parietal cortex and a commercial recording chamber (Crist Instruments, Hagerstown, MD) was implanted. The chamber was 20 mm in diameter and was attached to the skull by a ring of bone cement anchored by 6 to 8 titanium screws evenly distributed around the craniotomy. The chamber was placed above the STS, allowing a dorsal approach to MT. The precise location and shape of the STS for each monkey was determined from T2-weighted magnetic resonance images (MRIs) obtained in 2- and 1.5-T GE magnets with a small surface coil. This procedure was described in detail previously (Bisley and Pasternak 2000; Rudolph and Pasternak 1999). Briefly, the monkeys were anesthetized with sodium pentobarbital (25 mg/kg intravenously), and placed in a specially constructed MRI-compatible stereotaxic frame. Coronal and horizontal scans were performed with the following parameters: TE/TR: 5,000/90 or 3,000/85; 1.5-mm-thick slices 0.2 mm apart; 256 x 256 array, field of view: 10 or 15 cm.
Electrophysiological recordings
MAPPING AND CHARACTERIZATION OF RECEPTIVE FIELDS. Single-unit activity was recorded through tungsten microelectrodes (1.5–5.0 M
; FHC). The electrode was inserted through a guide tube positioned in a grid (Crist et al. 1988). Once a neuron was identified and well isolated, the position and size of its receptive field was mapped first by hand with a patch of random dots while the monkey passively viewed a fixation stimulus (a small cross) for variable periods of time ranging from 2 to 8 s. The optimal speed, dot density, and preferred direction were then determined with coherently moving random dots by selecting the parameters that produced maximal firing rates. A direction-selectivity profile was then generated by moving the random-dot stimulus at an optimal speed and with optimal dot density at least 5 times in each of the 4 cardinal and 4 oblique directions. Responses to each of these 8 directions were used to compute a vector average and the resulting direction was taken as the preferred direction of the cell. In about one third of the behavioral testing sessions, the sample stimulus could move in the preferred, the antipreferred (directly opposite to the preferred), or either of the 2 orthogonal directions. In the remaining sessions, possible sample directions were limited to the preferred and the antipreferred directions. Here, we examine only the data collected from the preferred and antipreferred directions. For each cell 8–30 trials were collected for each mean direction and for each direction range.
BASELINE ACTIVITY. Baseline activity was measured on each trial during the last 500–800 ms of the period of fixation preceding the presentation of the sample stimulus. It was averaged and used to evaluate neuronal activity (see example neurons in Fig. 2).
|
Action potentials during the trial were discriminated using a BAK Dual Window Discriminator and pulses were time stamped and stored together with information about the current stimulus using custommade software. Only well-isolated single units with an absolute refractory period were used for analysis that was performed using MATLAB (Mathworks).
For all statistical tests the firing rates recorded during the trial were compared with the baseline rates recorded throughout the session. Activity recorded during each testing session was visualized by convolving the raw data into spike density functions (SDFs) with a sigma of 15 ms (Richmond et al. 1987). These SDFs were used for visual inspection only and were not used in the subsequent analysis of firing rates. The rate of activity at different stages of the task was analyzed by computing the mean number of action potentials over a given epoch in repeated presentations. For stimuli with a net direction of motion, only correct trials were used in analysis unless otherwise stated. For 360°-direction range stimuli, all trials were included in analysis.
RESPONSES DURING SAMPLE AND TEST STIMULI. Responses to the visual stimuli were computed by averaging the activity during a 400-ms epoch of stimulus presentation that excluded the initial period of visual latency. The latency was calculated for each neuron, as follows. A threshold level was determined by computing the mean + 2SDs of the baseline activity and then sliding a 50-ms window in 1-ms steps along the spike train starting at stimulus onset. The response latency was defined as the midpoint of the 50-ms epoch in which the mean firing rate reached the threshold. For each neuron, a conventional direction selectivity (DS) index was computed on the basis of the response to the random-dot stimuli coherently moving in the preferred and antipreferred directions
 |
ROC ANALYSIS. To determine whether activity during the delay contains information about the direction of the preceding sample, we used a receiver operating characteristic (ROC) based analysis (Britten et al. 1992). This analysis computes the probability that an ideal observer could report the direction of motion in the sample based solely on the activity during the delay. We will refer to this as the direction discrimination probability (DDP). We performed this analysis on consecutive 100-ms epochs. For each epoch an ROC curve was created by setting 12 threshold levels of activity covering the range of firing rates that followed the 2 directions of motion. For each threshold level the probability that the sample stimulus had moved in the preferred or antipreferred direction was calculated. To this end, we asked what proportion of the trials for each direction showed activity greater than the threshold. These data were plotted creating an ROC curve, with the area under the curve representing the DDP. A DDP of 0.5 indicates no difference in the distribution of responses in that epoch following the 2 directions. A value of 1 indicates that the activity following the preferred sample was always higher than the highest activity following the antipreferred sample, and a value of 0 indicates that the activity following the preferred sample was always lower than the lowest activity following the antipreferred sample.
To test the significance of each DDP value we ran a permutation test. This was done by randomly distributing all the trials from a single epoch for a single neuron into 2 groups, independent of the actual sample direction. These groups were nominally called the preferred group and antipreferred group and contained the same number of trials as the experimentally obtained groups. The DDP was calculated from the redistributed data, and the procedure was repeated 2,000 times, creating a distribution of DDPs. We then determined where in the distribution the actual DDP lay (the P value). Values in the top or bottom 2.5% were defined as significant (i.e., P < 0.05, 2-tailed test).
ANALYSIS OF ACTIVITY DURING THE DELAY. Analysis of delay activity was performed by dividing the delay into 4 epochs: 200–400, 600–800, 1,000–1,200, and 1,300–1,500 ms and calculating the mean number of action potentials that occurred during each of the epochs.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONDISCLOSURESACKNOWLEDGMENTSREFERENCES |
|---|
; Maunsell and Van Essen 1983). The receptive field sizes increased with eccentricity from about 2° near the fovea to 10° at 23° eccentricity. Linear regression through all the data gave a slope of 0.44 with an intercept of 1.76 (R2 = 0.54). These properties are consistent with those reported previously in MT cells (e.g., Albright and Desimone 1987; Desimone and Ungerleider 1986). Figure 2 shows activity from 3 example neurons. Trials in which the preferred (left plots) or antipreferred (right plots) directions were presented during the sample are shown separately. In these examples all trials for each sample direction are shown, and for that reason the activity seen during the test period of the task reflects responses to both the preferred and antipreferred directions.
As is true of most MT neurons, responses were strong to a coherent stimulus moving in the preferred direction (sample responses in left plots), and low or suppressed when the antipreferred stimulus was presented (sample responses in right plots). The activity of these neurons recorded during the delay is representative of many neurons in our population. They show a short period of increased activity in the first third of the delay. Often this response was composed of a discrete burst of activity (Fig. 2A, right plot; Fig. 2C, left plot). After this early activity, many neurons showed a period of suppression during the middle of the delay and a slight increment in response toward the end. A striking feature of the activity recorded during the delay is the difference in the strength of the early delay activity following the preferred and antipreferred directions; often being greater following the antipreferred than the preferred direction.
Responses during stimulus presentation
SAMPLE: DIRECTION RANGE. Figure 3 illustrates responses of MT neurons to stimuli containing a range of local directions. The average stimulus response functions for the 162 neurons tested with the 4 standard direction ranges (0, 150, 300, and 360° range) are shown for stimuli moving in the preferred (solid circles) and antipreferred (open circles) directions. As the direction range in the sample increased, responses to the sample stimulus moving in the preferred direction decreased and responses to the sample moving in the antipreferred direction increased. Note that the neurons retained strong directional selectivity over a broad range of local directions, which decreased only when the direction range approached the level of a typical psychophysical threshold (i.e., over 320°; Bisley and Pasternak 2000).
|
We also assessed whether the response to the test was affected by the direction of the preceding stimulus. This was done by comparing responses to the test stimulus from trials in which the test direction either matched or did not match the direction of the sample. We found that following the preferred sample, responses to the test moving in the preferred direction were slightly, but significantly, smaller than after the antipreferred sample (means: 96.3 and 98.1 spikes/s; P < 0.05, paired t-test). This was also the case for the antipreferred test; it was slightly lower following preferred direction than following the antipreferred direction (means: 27.9 and 28.7 spikes/s; P < 0.05, paired t-test). The response to the antipreferred test following the preferred sample was also significantly lower than the response to the antipreferred sample (P < 0.05, paired t-test), suggesting that this effect is likely to represent a decrease in the neuronal response, rather than an increase related to the match between the directions of the test and sample stimuli. This effect is consistent with a weak adaptation effect (Van Wezel and Britten 2002).
Activity of MT neurons during the delay
PATTERN OF DELAY ACTIVITY. The examples in Fig. 2 illustrate the behavior of many MT neurons. They show that after the offset of the response to the sample activity did not return to baseline and remain there until the presentation of the test. Rather, it displayed a characteristic pattern with an early period of activation, followed by suppression and then by a subtle increase in firing shortly before the onset of the test stimulus. To examine the incidence of these features, we divided the delay into 4 equal epochs: 200–400, 600–800, 1,000–1,200, and 1,300–1,500 ms. The first 200 ms of the delay were excluded from the analysis, based on the observation that responses of MT neurons to visual stimuli often persist for about 100–150 ms after the stimulus (Britten and Heuer 1999). The epoch of 200–400 ms was chosen because it includes most of the activity bursts seen early in the delay. The second and third epochs were chosen primarily to investigate the suppression during the middle of the delay. The last epoch, the period immediately preceding the onset of the test stimulus, was chosen to examine the apparent increase in firing rate in a large proportion of neurons.
Figure 4 compares the firing rates of individual neurons to the baseline activity recorded during the fixation period, in the 500–800 ms preceding the sample. Activity following the preferred (left plots) and antipreferred (right plots) samples is plotted separately for each of the 4 epochs. The plots show that the activity of many neurons deviate from baseline with points scattered both above and below the unity line. During the early epoch (200–400 ms) the majority of cells showed excitation with firing rates above the baseline, whereas in the middle of the delay (600–800 and 1,000–1,200 ms) the activity of many neurons was suppressed, dropping below baseline. Toward the end of the delay, the population returned to a slightly more excited state. Each of these patterns was significant at the population level (P < 0.005, paired t-test).
|
|
2 test) and suggests that the excitation and suppression occurring in different epochs of the delay are not completely independent. We found that toward the end of the delay many neurons showed an increase in activity (see examples in Fig. 2). The data in Fig. 5 show that this increase was relatively common in the population of our neurons and there was an increase in the incidence of excitation toward the end of the delay. We quantified this increase by comparing the firing rates of all neurons from the third epoch (1,000–1,200 ms) to the rates measured during the final epoch for each direction (Fig. 6). Most neurons showed an increase in activity from the middle to the end of the delay and, for the population as a whole, this effect was highly significant (P << 0.001, paired t-test for each plot).
|
|
|
Representation of stimulus features in the delay activity
Thus far, we have shown that many neurons in MT increase or decrease their firing during the delay. In the remaining sections of the study, we will show that information about the preceding sample stimulus is contained in a subset of MT neurons that display the activity described above.
DIRECTIONALITY OF DELAY ACTIVITY. We examined whether activity during the delay reflects the direction of the preceding sample by comparing firing rates of individual neurons following the preferred and antipreferred direction. This comparison was performed on activity recorded during each of the 4 epochs representing early, middle, and late delay (Fig. 9). Only activity following the coherently moving sample was used in this analysis (0°-direction range). Early in the delay (200–400 ms), 24% of the neurons showed significantly higher firing rates following the antipreferred stimulus than the preferred stimulus (P < 0.0125, Bonferroni corrected t-test; circles), whereas only 4% showed the opposite effect (triangles). The rest of the neurons showed no significant directional bias in their activity (crosses). An analysis of the population activity revealed a significant shift toward higher firing rates following the antipreferred direction (P << 0.001, paired t-test). The directional bias of activity during the delay was also present during the second epoch (600–800 ms) despite the fact that this period of the delay was dominated by suppression (see Fig. 5). Although the number of neurons showing significant differences in firing rates decreased to 4%, the population continued showing a significant shift toward higher firing rates following the antipreferred direction (P < 0.01). In the later epochs (1,000–1,200 ms and the last 200 ms) the proportion of cells with significantly different activity following the 2 directions dropped to <3% and the population no longer appeared to show significant directional bias (P > 0.2).
|
|
|
|
ANALYSIS OF ERROR TRIALS. To determine whether the directional bias of delay activity shown in Fig. 10 was behaviorally relevant, we performed an ROC analysis for trials leading to errors. Because most of the errors occurred on trials in which the sample contained a direction range of 300°, the analysis was limited to this stimulus condition. Sixty-seven neurons had a sufficient number of error and correct trials (minimum of 2 per category per direction) for this analysis and the DDP values were computed every 100 ms beginning 500 ms before the start of the delay. The results of this analysis are shown in Fig. 13. Despite some variability in the data, most likely attributable to the relatively small number of trials, there were differences between the indices computed for the 2 types of trials. During the sample, the DDP values were dominated by the preferred direction and the values for the correct trials (solid circles) were consistently higher than those for the error trials (P < 0.04; 2-tailed paired t-test), suggesting that during error trials sample responses were slightly less directional. The opposite pattern was seen during the delay. The DDP values from both correct and error trials dropped below 0.5, indicating higher firing rates following the antipreferred direction. However, the values for correct trials were still consistently higher than those for the errors (P < 0.01; 2-tailed paired t-test), suggesting that on error trials the activity following the antipreferred direction was greater (or that the activity following the preferred direction was weaker). Furthermore, a comparison of the values measured for the correct trials with 300°-range sample with those for 0°-range sample revealed a striking similarity (gray line), suggesting that the values from the incorrect trials may have been abnormally low. Thus the greater directionality on error trials may be indicative of neuronal activity dominated by the factors not related to remembering sample direction (e.g., sensory adaptation).
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONDISCLOSURESACKNOWLEDGMENTSREFERENCES |
|---|
Responses to the direction range of random-dot stimuli
The responses of MT neurons to random-dot stimuli have been studied extensively (e.g., Britten et al. 1993; Newsome et al. 1989); however, the stimuli in those studies were composed of random motion with a certain percentage of the dots moving in the same direction. Our stimuli are quite different; they consist of local vectors representing a limited range of directions that must be integrated into a percept of global motion. These stimuli, introduced by Williams and Sekuler (1984), have been used extensively in psychophysical studies examining the properties of motion integration mechanisms in humans and animals (Pasternak et al. 1990; Rudolph and Pasternak 1999; Watamaniuk and Sekuler 1992; Watamaniuk et al. 1989; Williams et al. 1991). Although involvement of MT neurons in processing of such stimuli has been suggested by lesion studies (Bisley and Pasternak 2000; Pasternak and Merigan 1994; Rudolph and Pasternak 1999), the response of MT neurons to these stimuli has not been previously studied.
The present results confirm the conclusions from the lesion studies that MT neurons play a role in integration of local motion signals. We found that MT neurons responded well to these stimuli and retained their direction selectivity over a broad range of local directions up to the levels when the range of local directions approached the level of a psychophysical threshold for the monkeys (see Fig. 3). This behavior is consistent with previous reports that MT neurons perform an averaging algorithm when presented with complex motion consisting of more than one local directional signal (Britten and Heuer 1999; Ferrera and Lisberger 1997; Recanzone et al. 1997).
Responses to sample and test stimuli
Although modulation of responses by the matching stimuli during the performance of match-to-sample tasks has been observed in neurons in inferotemporal and prefrontal cortex (Miller et al. 1993), we did not find this type of activity in MT. However, we did see subtle effects of the preferred sample on the responses to the test stimuli occurring 1.5 s later. Specifically, responses to the preferred and antipreferred test were slightly weaker following the preferred sample than following the antipreferred sample. This observation is consistent with previous studies of stimulus adaptation in MT (Kohn and Movshon 2003; Petersen et al. 1985; Van Wezel and Britten 2002), although the magnitude of the effect is less in the present study. There are 3 possible reasons for the weaker effect. First, the sample stimulus was presented for only 500 ms, a much shorter adapting time than that used in the other studies. Second, the 1.5-s delay was substantially longer that the interstimulus time employed in the studies of adaptation in MT. Finally, it is possible that during a behavioral task like ours adaptation-induced changes in firing rates are less detectable because other task-related activity also affects neuronal firing rates. This final point is addressed in more detail below.
Activity during the delay
Area MT is known to play a key role in processing of complex motion and only recently has evidence emerged that its neurons may also make a contribution to remembering motion stimuli. This evidence was provided by studies of monkeys with MT lesions (Bisley and Pasternak 2000) and by applying microstimulation to MT (Bisley et al. 2001) during the performance of a task identical to that used in this study. Additional evidence has been provided by psychophysical experiments (Pasternak and Zaksas 2003; Zaksas et al. 2001).
The present data provide new insights into the nature of the participation of MT neurons suggested by these studies. Although we did not observe the type of sustained delay activity commonly found in cortical areas implicated in working memory (Funahashi et al. 1989; Fuster 1973; Miyashita and Chang 1988; Quintana and Fuster 1999; Rainer et al. 1998), we found that the activity of many neurons early in the delay contain information about the preceding stimulus, and that throughout most of the delay, a directional signal is present.
CAN SENSORY ADAPTATION ACCOUNT FOR DELAY ACTIVITY? Before discussing the relationship between the activity during the delay and storage of visual motion information, we must consider the possibility that the directionality present in the population may be caused by poststimulus sensory adaptation. Of the 5 studies that have examined motion adaptation in MT neurons (Kohn and Movshon 2003; Lisberger and Movshon 1999; Petersen et al. 1985; Priebe et al. 2002; Van Wezel and Britten 2002), none has specifically examined the activity during the interval that immediately followed the adapting stimulus. The only available information about the activity during the interstimulus interval comes from Britten (personal communication), who recently examined the activity of 68 of the neurons described in the study by Van Wezel and Britten (2002) during the 500 ms separating the 3-s adapting stimulus and the test (Britten, unpublished results). Although many of the features of the activity during the delay were similar to those we observed, there were some notable differences. The main similarity was the presence of excitation following the antipreferred direction in the early delay (200–400 ms) in both sets of data. A direct comparison of this activity revealed a small difference (P < 0.05, ANOVA), with our data showing slightly higher firing rates during early delay (mean firing rates after subtracting baseline activity: 7.3 vs. 4.5 spikes/s). The difference between the 2 data sets was more pronounced in the early delay following the preferred stimulus. Although most of the neurons in Britten's adaptation study were either suppressed or at baseline (mean firing rate: –0.17 spikes/s), activity of the majority of our neurons was enhanced (see Fig. 4, top left; mean firing rate: 4.7 spikes/s). This difference was highly significant (P < 0.0001) and implies that active engagement in the task may generate a bias in activity toward the preferred direction, which may balance the activity attributed to adaptation. This interpretation would explain the stronger directional bias we see on error trials (see Fig. 13)—when the monkey was performing poorly, adaptation may have had more impact than on trials in which the animal performed the task correctly. We should note that we cannot rule out the possibility that the differences in activity following the preferred direction are attributed to the different stimulus durations used in the 2 studies (3-s adapting stimulus compared with a 500-ms sample stimulus). However, it seems unlikely that a shorter stimulus would result in excitation while a longer stimulus would result in suppression. We also cannot rule out the possibility that a delayed burst after the presentation of a stimulus is characteristic of some MT neurons, and that this burst may be suppressed after a long adapting stimulus. If this was the case, then the activity we have observed is not necessarily task-dependent, although as we argue below, this does not preclude its being used to perform the task optimally.
EARLY DELAY ACTIVITY. The most interesting activity seen in MT during the delay was found in the first 500 ms when a majority of neurons had responses significantly higher than baseline following both directions. We were careful to discount any response that could be directly attributed to the sample by ignoring the activity in the first 200 ms of the delay. Although in many neurons early delay activity was directionally biased, in a large number of cells this activity showed no directionality, suggesting the presence of a nondirectional input into MT. It is tempting to speculate that the source of this activity may be an area involved in working memory. However, it is also possible that the nondirectional signal originates from within MT and is a neural correlate of the normalization process suggested by models of MT (Koechlin et al. 1999; Simoncelli and Heeger 1998).
One third of the neurons in MT showed significant directionality early in the delay. A small number of them had higher firing rates following the preferred direction and this activity was affected only by the increase of the direction range of the preferred sample. Broadening of the direction range of the antipreferred sample had no affect on their firing rate. The remaining neurons, representing 29% of the entire population recorded, showed the opposite effect: their activity was greater following the antipreferred sample and was strongly affected by the direction range in that stimulus but was unaffected by the direction range of the preferred sample. These data show that activity early in the delay contains information about both the direction and the range of motion in the sample stimulus, although it appears to be segregated by sample direction— one population encoding changes in the preferred and the other encoding changes in the antipreferred direction. The existence of separate subpopulations of neurons showing opposing types of activity during the delay may be an important feature of the network underlying memory and may offer insights as to how the stimulus features are stored.
There are 2 independent lines of evidence that suggest that the delay activity we see in MT may be involved in the memory process. First, animals with MT lesions showed a loss in stored directional information that cannot be explained by the deficit in encoding alone (Bisley and Pasternak 2000). Second, the presentation of a noisy mask early in the delay is maximally effective at disrupting the memory of a previously seen directional stimulus (Pasternak and Zaksas 2003). Together these data suggest that activity in MT following the remembered sample, whether it is attributed to sensory adaptation or represents a more active process, may contribute to the network underlying memory for motion.
We observed that neurons with higher early activity following the antipreferred sample also tended to have significantly higher maximal firing rates than neurons with early activation that was not directionally biased. It should be pointed out these neurons were selected on the basis of their activity during the delay and the nature of the directional bias. The finding that these neurons also share another characteristic— high maximal firing rates to the preferred stimuli—suggests that they are likely to represent a separate class of MT neurons.
SUPPRESSION DURING THE MIDDLE OF THE DELAY. About half of the neurons showed significant suppression during the middle of the delay following either the preferred or antipreferred stimulus, and in many cases this suppression was greater following the preferred direction. It should be pointed out that a decrease in activity following the preferred adapting stimulus was observed during the test by Van Wezel and Britten (2002) as well as during the 500-ms interstimulus interval (Britten, personal communication). Thus some of the suppression following the preferred direction could in part be related to poststimulus adaptation processes. On the other hand, it is less likely that the prolonged suppression following the antipreferred sample is also attributable to the adaptation, given that no suppression was observed by Britten, who would be more likely to find such effects given the much longer adapting stimulus than the 500-ms sample we used. Thus it is more likely that the suppression we observed during middle delay is largely task-related.
Sustained suppression during the delay is not unique to MT and was previously observed in other cortical areas, including somatosensory regions of parietal cortex (Koch and Fuster 1989; Zhou and Fuster 1996) and extrastriate area V3a (Nakamura and Colby 2000). Although the role of this suppression is not clear, it may limit the amount of irrelevant sensory information reaching cortical areas involved in active retention of the preceding sample, and could therefore serve to decrease the background noise and to enhance neural responses to the anticipated test stimulus (Koch and Fuster 1989). A similar gating mechanism downstream of MT has been proposed in a task in which monkeys had to delay a motor response to a motion stimulus (Seidemann et al. 1998). Although the area in which the gating occurs may differ in the 2 tasks, the suppression we see during the delay may represent a neuronal correlate of this type of mechanism.
REACTIVATION LATE IN THE DELAY. In most of the testing sessions the delay duration was 1,500 ms. Thus the end of the delay and the appearance of the test were highly predictable. This predictability was reflected in many neurons in the form of increased activity during the last several hundred milliseconds of the delay. When the delay was increased to 3,000 ms, the same increase in activity was seen by the end of the delay, but the increase was more gradual and began at about the time when the more commonly encountered 1,500-ms delay trials would have ended. If the test stimulus was replaced by a reward for a block of trials, the increase in activity was no longer found. These observations strongly point to the anticipatory nature of this activity, and most likely represent "top-down" signals originating elsewhere in the brain.
A number of studies have reported an increase in activity in visual neurons before the appearance of behaviorally relevant stimuli. This effect has been observed in several extrastriate cortical areas (Colby et al. 1996; Luck et al. 1997; Nakamura and Colby 2000; Recanzone and Wurtz 2000) and has been attributed to directing spatial attention to the receptive field. Thus it is possible that this reactivation may be the result of a shift of spatial attention to the location of the upcoming test. Indeed, a number of studies have demonstrated that MT neurons are affected by the attentional demands of the task (Cook and Maunsell 2002; Martinez-Trujillo and Treue 2002; Seidemann and Newsome 1999; Treue and Martinez Trujillo 1999; Treue and Maunsell 1999). However, it is also possible that other cognitive influences, such as anticipation, may also contribute to this effect. Indeed such activity has been observed in prefrontal cortex during behavioral tasks resembling the task used here (Boch and Goldberg 1989; Niki and Watanabe 1979; Quintana and Fuster 1999; Rainer and Miller 2002; Rainer et al. 1999; Romo et al. 1999).
Toward the end of the delay, the directional bias of the population became quite weak. This apparent absence of directionality may be related to the presence of the anticipatory signal during this period. It is possible that the mechanism that produces this anticipatory activity is independent of the mechanism involved in adaptation and the signals it generates may be strong enough to make any residual directional bias difficult to detect. However, we cannot rule out the possibility that the decline in directionality seen in Fig. 10 at least in part represents the time course of the effects of adaptation, and the very weak adaptive effect we saw on the test stimuli is because the adaptation effect is so weak by the end of the delay.
Comparison with other studies
Only 2 previous studies have examined the behavior of MT neurons during the performance of tasks involving working memory. Seidemann et al. (1998) trained monkeys in a task requiring them to withhold a saccade indicating the direction of the preceding motion stimulus. Thus the monkey was required to remember the location to which the saccade would eventually be made, rather than the direction of stimulus motion. They noted, referring to unpublished observations, that recordings from MT neurons during the poststimulus delay revealed no significant activity.
Ferrera et al. (1994) recorded from MT using a task in which the monkeys were required to remember the direction of visual motion. In that study, a random-dot motion stimulus served as a cue and, after a variable delay of 200–540 ms, was followed by a sequence of stimuli one of which matched the direction of the cue. The monkeys responded by releasing a bar to the matching stimulus. The analysis was limited to a maximal period of about 200 ms, beginning 350 ms after the offset of the cue. About a third of MT neurons had substantially elevated activity, although some showed suppression during this period. Because there was little correlation between this activity and the identity of the cue, it was concluded that the delay activity was unlikely to carry information about the preceding stimulus. Apart from task differences, 2 other factors should be considered when comparing these findings to our results: differences in the delay length and in the delay epochs that were analyzed. The period between 350 and 540 ms in Ferrera et al. (1994) represents a time when most of the early delay activation we observed would already have occurred and the difference in early activation for the preferred and the antipreferred directions would have been missed. On the other hand, the significant excitation and suppression observed by these authors during the delay may correspond to some of the activity we observed during that same period. Thus there are substantial similarities between our results and those reported by Ferrera et al. (1994) and the differences are likely to be attributable largely to the different delay lengths and the phase of the delay chosen for analysis.
Match-to-sample tasks, similar to the task used here, have also been used in studies of neurons in prefrontal (Fuster and Alexander 1971; Miller et al. 1996; Rainer and Miller 2002), parietal (Sereno and Maunsell 1998), and inferotemporal (Fuster and Jervey 1981) cortices. Many neurons in these areas display sustained elevation of activity during the delay that appears to carry the information about the preceding stimulus. This type of persistent activity is commonly thought to play an important role of bridging the interval between the sensory stimulus and the behavioral response based on that stimulus (Fuster 1995b). Most of these studies emphasize the sustained nature of the delay activity and less attention has been paid to its temporal dynamics. However, recently, Rainer and Miller (2002) performed such an analysis and identified 3 activity periods, some aspects of which resembled the delay activity we observed. During the stimulus, prefrontal neurons showed a transient response followed by a burst of activity reminiscent of the activity we found early in the delay in MT. This "intermediate" period was followed by a period of "reactivation" during which activity continued to increase until the second stimulus was presented. It is noteworthy that middle delay suppression, common in area MT and in other sensory cortical areas, did not appear to be present in prefrontal neurons.
The dynamics of delay activity in prefrontal neurons in response to vibrotactile stimuli is also similar to some features of the delay activity observed in this study in MT. Romo et al. (1999) reported that many neurons fired in proportion to the frequency of the remembered stimulus, a result analogous to the relationship between firing during early delay and direction range. These authors also noted the presence of early delay activity in secondary somatosensory cortex (Salinas et al. 1998).
Are MT neurons involved in temporary storage of motion information?
The notion that the same systems involved in sensory processing also participate in retaining sensory information is generally accepted and incorporated into most models of working memory (e.g., Fuster 1997; Goldman-Rakic 1995). It is based on experiments showing that during the memory period firing rates of neurons in areas processing sensory signals reflect the identity of the remembered stimulus (Fuster and Jervey 1982; Miyashita and Chang 1988; Salinas et al. 1998; Zhou and Fuster 1996). This notion is also supported by recent lesion and microstimulation studies (Bisley and Pasternak 2000; Bisley et al. 2001), as well as by functional imaging in humans (Courtney et al. 1997; Haxby et al. 2000).
Prefrontal cortex is the region most closely associated with working memory (Fuster 1995a; Goldman-Rakic 1995; Miller et al. 1996). It receives direct projections from MT (Barbas 1988; Schall et al. 1995) and sends direct projections back (Cusick et al. 1995). These connections provide the basis for the functional interaction between the 2 regions. Indeed, neurons in dorsolateral prefrontal cortex respond to moving random-dot stimuli and their firing rates appear to reflect stimulus strength and the direction of motion (Kim and Shadlen 1999). Although the activity of prefrontal neurons during memory tasks requiring remembering visual motion has not been examined, the similarity between the behaviors of prefrontal and MT neurons during the delay suggests that both areas may participate in the same circuitry underlying the retention of motion information.
Our findings support the notion that MT may be involved in the circuitry underlying storage of motion information. There is a resurgence of activity, some of which is strongly related to the preceding stimulus, early in the delay, and a weak directional signal that continues throughout the delay. Some aspects of this activity are likely to be related to sensory adaptation to the preceding stimulus, particularly following the antipreferred sample. However, other aspects of this activity cannot be explained by passive adaptation alone. It remains to be seen whether the dynamics of this delay activity is specific to MT neurons or is present in directionally selective neurons in other cortical areas and in what way it reflects the workings of the network underlying the ability to remember visual motion.
|
DISCLOSURES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONDISCLOSURESACKNOWLEDGMENTSREFERENCES |
|---|
|
ACKNOWLEDGMENTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONDISCLOSURESACKNOWLEDGMENTSREFERENCES |
|---|
Present address of J. W. Bisley: Mahoney Center for Brain and Behavior, Center for Neurobiology and Behavior, Columbia University, New York, NY 10027.
|
FOOTNOTES |
|---|
Address for reprint requests and other correspondence: T. Pasternak, Department of Neurobiology and Anatomy, Box 603, University of Rochester, Rochester, NY 14642 (E-mail: tania{at}cvs.rochester.edu).
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONDISCLOSURESACKNOWLEDGMENTSREFERENCES |
|---|
Asaad WF, Rainer G, and Miller EK. Task-specific neural activity in the primate prefrontal cortex. J Neurophysiol 84: 451–459, 2000.
Barbas H. Anatomic organization of basoventral and mediodorsal visual recipient prefrontal regions in the rhesus monkey. J Comp Neurol 276: 313–342, 1988.[CrossRef][Web of Science][Medline]
Bisley JW and Pasternak T. The multiple roles of visual cortical areas MT/MST in remembering the direction of visual motion. Cereb Cortex 10: 1053–1065, 2000.
Bisley JW, Zaksas D, and Pasternak T. Microstimulation of cortical area MT affects performance on a visual working memory task. J Neurophysiol 85: 187–196, 2001.
Boch RA and Goldberg ME. Participation of prefrontal neurons in the preparation of visually guided eye movements in the rhesus monkey. J Neurophysiol 61: 1064–1084, 1989.
Britten KH and Heuer HW. Spatial summation in the receptive fields of MT neurons. J Neurosci 19: 5074–5084, 1999.
Britten KH, Newsome WT, Shadlen MN, Celebrini S, and Movshon JA. A relationship between behavioral choice and the visual responses of neurons in macaque MT. Vis Neurosci 13: 87–100, 1996.[Web of Science][Medline]
Britten KH, Shadlen MN, Newsome WT, and Movshon JA. The analysis of visual motion: a comparison of neuronal and psychophysical performance. J Neurosci 12: 4745–4765, 1992.[Abstract]
Britten KH, Shadlen MN, Newsome WT, and Movshon JA. Responses of neurons in macaque MT to stochastic motion signals. Vis Neurosci 10: 1157–1169, 1993.[Web of Science][Medline]
Colby CL, Duhamel JR, and Goldberg ME. Visual, presaccadic, and cognitive activation of single neurons in monkey lateral intraparietal area. J Neurophysiol 76: 2841–2852, 1996.
Cook EP and Maunsell JH. Attentional modulation of behavioral performance and neuronal responses in middle temporal and ventral intraparietal areas of macaque monkey. J Neurosci 22: 1994–2004, 2002.
Courtney SM, Ungerleider BG, Keil K, and Haxby JV. Transient and sustained activity in a distributed neural system for human working memory. Nature 386: 608–611, 1997.[CrossRef][Medline]
Crist CF, Yamasaki DSG, Komatsu H, and Wurtz RH. A grid system and a microsyringe for single cell recording. J Neurosci Methods 26: 117–122, 1988.[CrossRef][Web of Science][Medline]
Cusick CG, Seltzer B, Cola M, and Griggs E. Chemoarchitectonics and corticocortical terminations within the superior temporal sulcus of the rhesus monkey: evidence for subdivisions of superior temporal polysensory cortex. J Comp Neurol 360: 513–535, 1995.[CrossRef][Web of Science][Medline]
Desimone R and Ungerleider LG. Multiple visual areas in the caudal superior temporal sulcus of the macaque. J Comp Neurol 248: 164–189, 1986.[CrossRef][Web of Science][Medline]
Ditterich J, Mazurek ME, and Shadlen MN. Microstimulation of visual cortex affects the speed of perceptual decisions. Nat Neurosci 6: 891–898, 2003.[CrossRef][Web of Science][Medline]
Ferrera VP and Lisberger SG. Neuronal responses in visual areas MT and MST during smooth pursuit target selection. J Neurophysiol 78: 1433–1446, 1997.
Ferrera VP, Rudolph KK, and Maunsell JH. Responses of neurons in the parietal and temporal visual pathways during a motion task. J Neurosci 14: 6171–6186, 1994.[Abstract]
Funahashi S, Bruce CJ, and Goldman-Rakic PS. Mnemonic coding of visual space in the monkey's dorsolateral prefrontal cortex. J Neurophysiol 61: 331–349, 1989.
Fuster JM. Unit activity in prefrontal cortex during delayed-response performance: neuronal correlates of transient memory. J Neurophysiol 36: 61–78, 1973.
Fuster JM. Memory in the Cerebral Cortex. Cambridge, MA: MIT Press, 1995a.
Fuster JM. Memory in the cortex of the primate. Biol Res 28: 59–72, 1995b.[Medline]
Fuster JM. Network memory. Trends Neurosci 20: 451–459, 1997.[CrossRef][Web of Science][Medline]
Fuster JM and Alexander GE. Neuron activity related to short-term memory. Science 173: 652–654, 1971.
Fuster JM and Jervey JP. Inferotemporal neurons distinguish and retain behaviorally relevant features of visual stimuli. Science 212: 952–955, 1981.
Fuster JM and Jervey JP. Neuronal firing in the inferotemporal cortex of the monkey in a visual memory task. J Neurosci 2: 361–375, 1982.[Abstract]
Goldman-Rakic PS. Cellular basis of working memory. Neuron 14: 477–485, 1995.[CrossRef][Web of Science][Medline]
Haxby JV, Petit L, Ungerleider LG, and Courtney SM. Distinguishing the functional roles of multiple regions in distributed neural systems for visual working memory. Neuroimage 11: 145–156, 2000.[CrossRef][Web of Science][Medline]
Kim JN and Shadlen MN. Neural correlates of a decision in the dorsolateral prefrontal cortex of the macaque. Nat Neurosci 2: 176–185, 1999.[CrossRef][Web of Science][Medline]
Koch KW and Fuster JM. Unit activity in monkey parietal cortex related to haptic perception and temporary memory. Exp Brain Res 76: 292–306, 1989.[Web of Science][Medline]
Koechlin E, Anton JL, and Burnod Y. Bayesian inference in populations of cortical neurons: a model of motion integration and segmentation in area MT. Biol Cybern 80: 25–44, 1999.[CrossRef][Web of Science][Medline]
Kohn A and Movshon JA. Neuronal adaptation to visual motion in area MT of the macaque. Neuron 39: 681–691, 2003.[CrossRef][Web of Science][Medline]
Konorski J. A new method of physiological investigation of recent memory in animals. Bull Acad Pol Sci Biol 7: 115–119, 1959.
Lisberger SG and Movshon JA. Visual motion analysis for pursuit eye movements in area MT of macaque monkeys. J Neurosci 19: 2224–2246, 1999.
Luck SJ, Chelazzi L, Hillyard SA, and Desimone R. Neural mechanisms of spatial selective attention in areas V1, V2, and V4 of macaque visual cortex. J Neurophysiol 77: 24–42, 1997.
Martinez-Trujillo J and Treue S. Attentional modulation strength in cortical area MT depends on stimulus contrast. Neuron 35: 365–370, 2002.[CrossRef][Web of Science][Medline]
Maunsell JH and Van Essen DC. Functional properties of neurons in middle temporal visual area of the macaque monkey. I. Selectivity for stimulus direction, speed, and orientation. J Neurophysiol 49: 1127–1147, 1983.
Merigan WH and Pasternak T. Lesions in primate visual cortex leading to deficits of perception. In: Neuropsychology of Vision, edited by Fahle M and Greenlee M. New York: Oxford Univ. Press, 2002.
Miller EK and Cohen JD. An integrative theory of prefrontal cortex function. Annu Rev Neurosci 24: 167–202, 2001.[CrossRef][Web of Science][Medline]
Miller EK and Desimone R. Parallel neuronal mechanisms for short-term memory. Science 263: 520–522, 1994.
Miller EK, Erickson CA, and Desimone R. Neural mechanisms of visual working memory in prefrontal cortex of the macaque. J Neurosci 16: 5154–5167, 1996.
Miller EK, Li L, and Desimone R. Activity of neurons in anterior inferior temporal cortex during a short-term memory task. J Neurosci 13: 1460–1478, 1993.[Abstract]
Miyashita Y and Chang HS. Neuronal correlate of pictorial short-term memory in the primate temporal cortex. Nature 331: 68–70, 1988.[CrossRef][Medline]
Nakamura K and Colby CL. Visual, saccade-related, and cognitive activation of single neurons in monkey extrastriate area V3a. J Neurophysiol 84: 677–692, 2000.
Newsome WT, Britten KH, and Movshon JA. Neuronal correlates of a perceptual decision. Nature 341: 52–54, 1989.[CrossRef][Medline]
Nichols MJ and Newsome WT. Middle temporal visual area microstimulation influences veridical judgments of motion direction. J Neurosci 22: 9530–9540, 2002.
Niki H and Watanabe M. Prefrontal and cingulate unit activity during timing behavior in the monkey. Brain Res 171: 213–224, 1979.[CrossRef][Web of Science][Medline]
Pasternak T, Albano JE, and Harvitt DM. The role of directionally selective neurons in the perception of global motion. J Neurosci 10: 3079–3086, 1990.[Abstract]
Pasternak T, Bisley JW, and Calkins D. Visual information processing in the primate brain. In: Biological Psychology, edited by Gallagher M and Nelson RJ. New York: Wiley, 2003, p. 130–185.
Pasternak T and Merigan WH. Motion perception following lesions of the superior temporal sulcus in the monkey. Cereb Cortex 4: 247–259, 1994.
Pasternak T and Zaksas D. Stimulus specificity and temporal dynamics of working memory for visual motion. J Neurophysiol 90: 2752–2757, 2003.
Petersen SE, Baker JF, and Allman JM. Direction-specific adaptation in area MT of the owl monkey. Brain Res 346: 146–150, 1985.[CrossRef][Web of Science][Medline]
Priebe NJ, Churchland MM, and Lisberger SG. Constraints on the source of short-term motion adaptation in macaque area MT. I. the role of input and intrinsic mechanisms. J Neurophysiol 88: 354–369, 2002.
Quintana J and Fuster JM. From perception to action: temporal integrative functions of prefrontal and parietal neurons. Cereb Cortex 9: 213–221, 1999.
Rainer G, Asaad WF, and Miller EK. Memory fields of neurons in the primate prefrontal cortex. Proc Natl Acad Sci USA 95: 15008–15013, 1998.
Rainer G and Miller EK. Timecourse of object-related neural activity in the primate prefrontal cortex during a short-term memory task. Eur J Neurosci 15: 1244–1254, 2002.[CrossRef][Web of Science][Medline]
Rainer G, Rao SC, and Miller EK. Prospective coding for objects in primate prefrontal cortex. J Neurosci 19: 5493–5505, 1999.
Recanzone GH and Wurtz RH. Effects of attention on MT and MST neuronal activity during pursuit initiation. J Neurophysiol 83: 777–790, 2000.
Recanzone GH, Wurtz RH, and Schwarz U. Responses of MT and MST neurons to one and two moving objects in the receptive field. J Neurophysiol 78: 2904–2915, 1997.
Remmel RS. An inexpensive eye movement monitor using the scieral search coil technique. IEEE Trans Biomed Eng 31: 388–390, 1984.[Web of Science][Medline]
Richmond BJ, Optican LM, Podell M, and Spitzer H. Temporal encoding of two-dimensional patterns by single units in primate inferior temporal cortex. I. Response characteristics. J Neurophysiol 57: 132–146, 1987.
Romo R, Brody CD, Hernandez A, and Lemus L. Neuronal correlates of parametric working memory in the prefrontal cortex. Nature 399: 470–473, 1999.[CrossRef][Medline]
Rudolph K and Pasternak T. Transient and permanent deficits in motion perception after lesions of cortical areas MT and MST in the macaque monkey. Cereb Cortex 9: 90–100, 1999.
Salinas E, Hernandez A, Zainos A, Lemus L, and Romo R. Cortical recoding of sensory stimuli during somatosensory discrimination. Soc Neurosci Abstr 24: 1126, 1998.
Salzman CD, Murasugi CM, Britten KH, and Newsome WT. Microstimulation in visual area MT: effects on direction discrimination performance. J Neurosci 12: 2331–2355, 1992.[Abstract]
Schall JD, Morel A, King DJ, and Bullier J. Topography of visual cortex connections with frontal eye field in macaque: convergence and segregation of processing streams. J Neurosci 15: 4464–4487, 1995.[Abstract]
Seidemann E and Newsome WT. Effect of spatial attention on the responses of area MT neurons. J Neurophysiol 81: 1783–1794, 1999.
Seidemann E, Zohary E, and Newsome WT. Temporal gating of neural signals during performance of a visual discrimination task. Nature 394: 72–75, 1998.[CrossRef][Medline]
Sereno AB and Maunsell JH. Shape selectivity in primate lateral intraparietal cortex. Nature 395: 500–503, 1998.[CrossRef][Medline]
Simoncelli EP and Heeger DJ. A model of neuronal responses in visual area MT. Vision Res 38: 743–761, 1998.[CrossRef][Web of Science][Medline]
Treue S and Martinez Trujillo JC. Feature-based attention influences motion processing gain in macaque visual cortex. Nature 399: 575–579, 1999.[CrossRef][Medline]
Treue S and Maunsell JHR. Effects of attention on the processing of motion in macaque middle temporal and medial superior temporal visual cortical areas. J Neurosci 19: 7591–7602, 1999.
Van Wezel RJA and Britten KH. Motion adaptation in area MT. J Neurophysiol 88: 3469–3476, 2002.
Watamaniuk SN and Sekuler R. Temporal and spatial integration in dynamic random-dot stimuli. Vision Res 32: 2341–2347, 1992.[CrossRef][Web of Science][Medline]
Watamaniuk SN, Sekuler R, and Williams DW. Direction perception in complex dynamic displays: the integration of direction information. Vision Res 29: 47–59, 1989.[CrossRef][Web of Science][Medline]
Williams D, Tweten S, and Sekuler R. Using metamers to explore motion perception. Vision Res 31: 275–286, 1991.[CrossRef][Web of Science][Medline]
Williams DW and Sekuler R. Coherent global motion percepts from stochastic local motions. Vision Res 24: 55–62, 1984.[CrossRef][Web of Science][Medline]
Zaksas D, Bisley JW, and Pasternak T. Motion information is spatially localized in a visual working-memory task. J Neurophysiol 86: 912–921, 2001.
Zhou YD and Fuster JM. Mnemonic neuronal activity in somatosensory cortex. Proc Natl Acad Sci USA 93: 10533–10537, 1996.
This article has been cited by other articles:
|
|
 |
|
  A. Mojzisch and K. Krug Cells, circuits, and choices: Social influences on perceptual decision making Cogn Affect Behav Neurosci, December 1, 2008; 8(4): 498 - 508. [Abstract] [PDF]
|
|
|
|
 |
|
 M. J. WENGER, A. M. COPELAND, J. L. BITTNER, and R. D. THOMAS Evidence for criterion shifts in visual perceptual learning: Data and implications Atten Percept Psychophys, October 1, 2008; 70(7): 1248 - 1273. [Abstract] [PDF]
|
|
|
|
 |
|
 M.R. Burke and G.R. Barnes Brain and Behavior: A Task-Dependent Eye Movement Study Cereb Cortex, January 1, 2008; 18(1): 126 - 135. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Ardid, X.-J. Wang, and A. Compte An Integrated Microcircuit Model of Attentional Processing in the Neocortex J. Neurosci., August 8, 2007; 27(32): 8486 - 8495. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. R. Lehky and K. Tanaka Enhancement of Object Representations in Primate Perirhinal Cortex During a Visual Working-Memory Task J Neurophysiol, February 1, 2007; 97(2): 1298 - 1310. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. Zaksas and T. Pasternak Directional Signals in the Prefrontal Cortex and in Area MT during a Working Memory for Visual Motion Task. J. Neurosci., November 8, 2006; 26(45): 11726 - 11742. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. Zaksas and T. Pasternak Area MT Neurons Respond to Visual Motion Distant From Their Receptive Fields J Neurophysiol, December 1, 2005; 94(6): 4156 - 4167. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Constantinidis and X.-J. Wang A Neural Circuit Basis for Spatial Working Memory Neuroscientist, December 1, 2004; 10(6): 553 - 565. [Abstract] [PDF]
|
|
|
|
|
|
S. J. Bennett and G. R. Barnes Predictive Smooth Ocular Pursuit During the Transient Disappearance of a Visual Target J Neurophysiol, July 1, 2004; 92(1): 578 - 590. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Visit Other APS Journals Online |
