A model of curved saccade trajectories: Spike rate adaptation in the brainstem as the cause of deviation away

The trajectory of saccades to a target is often affected whenever there is a distractor in the visual field. Distractors can cause a saccade to deviate towards their location or away from it. The oculomotor mechanisms that produce deviation towards distractors have been thoroughly explored in behavioral, neurophysiological and computational studies. The mechanisms underlying deviation away, on the other hand, remain unclear. Behavioral findings suggest a mechanism of spatially focused, top-down inhibition in a saccade map, and deviation away has become a tool to investigate such inhibition. However, this inhibition hypothesis has little neuroanatomical or neurophysiological support, and recent findings go against it. Here, we propose that deviation away results from an unbalanced saccade drive from the brainstem, caused by spike rate adaptation in brainstem long-lead burst neurons. Adaptation to stimulation in the direction of the distractor results in an unbalanced drive away from it. An existing model of the saccade system was extended with this theory. The resulting model simulates a wide range of findings on saccade trajectories, including findings that have classically been interpreted to support inhibition views. Furthermore, the model replicated the effect of saccade latency on deviation away, but predicted this effect would be absent with large (400 ms) distractor-target onset asynchrony. This prediction was confirmed in an experiment, which demonstrates that the theory both explains classical findings on saccade trajectories and predicts new findings.     

Gaze and arrow distractors influence saccade trajectories similarly

Perceiving someone's averted eye-gaze is thought to result in an automatic shift of attention and in the preparation of an oculomotor response in the direction of perceived gaze. Although gaze cues have been regarded as being special in this respect, recent studies have found evidence for automatic attention shifts with nonsocial stimuli, such as arrow cues. Here, we directly compared the effects of social and nonsocial cues on eye movement preparation by examining the modulation of saccade trajectories made in the presence of eye-gaze, arrows, or peripheral distractors. At a short stimulus onset asynchrony (SOA) between the distractor and the target, saccades deviated towards the direction of centrally presented arrow distractors, but away from the peripheral distractors. No significant trajectory deviations were found for gaze distractors. At the longer SOA, saccades deviated away from the direction of the distractor for all three distractor types, but deviations were smaller for the centrally presented gaze and arrow distractors. These effects were independent of whether line-drawings or photos of faces were used and could not be explained by differences in the spatial properties of the peripheral distractor. The results suggest that all three types of distractors (gaze, arrow, peripheral) can induce the automatic programming of an eye movement. Moreover, the findings suggest that gaze and arrow distractors affect oculomotor preparation similarly, whereas peripheral distractors, which are classically regarded as eliciting an automatic shift of attention and an oculomotor response, induce a stronger and faster acting influence on response preparation and the corresponding inhibition of that response.     

The effects of response priming on the planning and execution of goal-directed movements in the presence of a distracting stimulus

Two models of selective reaching have been proposed to account for deviations in movement trajectories in cluttered environments. The response vector model predicts movement trajectories should deviate towards or away from the location a distractor of little or large salience, respectively. In contrast, the response activation model predicts that a distractor with large salience should cause movement deviations towards it whereas a distractor with little salience should not influence the movement. The precuing technique was combined with the distractor interference paradigm to test these predictions. Results indicate that when the target was presented at the precued (salient) location, movements were unaffected by a distractor. Conversely, when the distractor was presented at the precued location while the target was presented at an uncued (non-salient) location, participants demonstrated increased reaction times and trajectory deviations towards the location of the distractor. These findings are consistent with the model of response activation.     

Attentional capture with various distractor and target types

Effects of nonpredictive distractors that involved changes in luminance, size, or shape were examined in three experiments. In Experiment 1, with two types of distractors (onsets and offsets), accuracy was better on trials when the distractor was near the location of either an offset or an onset target than on trials when the distractor was in a different location from that of the target, demonstrating attentional capture. Capture occurred both when the type of target (onset or offset) was blocked and therefore predictable and also when the type of target was mixed within blocks and therefore not predictable. Further experiments indicated that distractors captured attention even when the change to distractor did not create a new perceptual object. Neither a singleton-detection mode, nor a contingent involuntary orienting hypothesis, nor creation of a new object seems to explain all of these data adequately. Rather, capture may depend on a number of factors in the task.     

Attentional capture with various distractor and target types

Effects of nonpredictive distractors that involved changes in luminance, size, or shape were examined in three experiments. In Experiment 1, with two types of distractors (onsets and offsets), accuracy was better on trials when the distractor was near the location of either an offset or an onset target than on trials when the distractor was in a different location from that of the target, demonstrating attentional capture. Capture occurred both when the type of target (onset or offset) was blocked and therefore predictable and also when the type of target was mixed within blocks and therefore not predictable. Further experiments indicated that distractors captured attention even when the change to distractor did not create a new perceptual object. Neither a singleton-detection mode, nor a contingent involuntary orienting hypothesis, nor creation of a new object seems to explain all of these data adequately. Rather, capture may depend on a number of factors in the task.     

Remembering a location makes the eyes curve away

Working memory is a system that keeps limited information on-line for immediate access by cognitive processes. This type of active maintenance is important for everyday life activities. The present study shows that maintaining a location in spatial working memory affects the trajectories of saccadic eye movements toward visual targets, as the eyes deviate away from the remembered location. This finding provides direct evidence for a strong overlap between spatial working memory and the eye movement system. We argue that curvature is the result of the need to inhibit memory-based eye movement activity in the superior colliculus, in order to allow an accurate saccade to the visual target. Whereas previous research has shown that the eyes may deviate away from visually presented stimuli that need to be ignored, we show that the eyes also curve away from remembered stimuli.     

The relationship between inhibition of return and saccade trajectory deviations

After presentation of a peripheral cue, a subsequent saccade to the cued location is delayed (inhibition of return: IOR). Furthermore, saccades typically deviate away from the cued location. The present study examined the relationship between these inhibitory effects. IOR and saccade trajectory deviations were found after central (endogenous) and peripheral (exogenous) cuing of attention, and both effects were larger with an onset cue than with a color singleton cue. However, a dissociation in time course was found between IOR and saccade trajectory deviations. Saccade trajectory deviations occurred at short delays between the cue and the saccade, but IOR was found at longer delays. A model is proposed in which IOR is caused by inhibition applied to a preoculomotor attentional map, whereas saccade trajectory deviations are caused by inhibition applied to the saccade map, in which the final stage of oculomotor programming takes place.     

Top-down control in time and space: Evidence from saccadic latencies and trajectories

Visual distractors disrupt the production of saccadic eye movements temporally, by increasing saccade latency, and spatially, by biasing the trajectory of the movement. The present research investigated the extent to which top-down control can be exerted over these two forms of oculomotor capture. In two experiments, people were instructed to make target directed saccades in the presence of distractors, and temporal and spatial capture were assessed simultaneously by measuring saccade latency and saccade trajectory curvature, respectively. In Experiment 1, an attentional control set manipulation was employed, resulting in the elimination of temporal capture, but only an attenuation of spatial capture. In Experiment 2, foreknowledge of the target location caused an attenuation of temporal capture but an enhancement of spatial capture. These results suggest that, whereas temporal capture is contingent on top-down control, the spatial component of capture is stimulus-driven.     

Grasping remaps the distribution of visuospatial attention and enhances competing action activation

We examined how action goals influence the distribution of visuospatial attention near the body (Experiment 1) and how the temporal relationship between distractors and targets modifies shifts in visuospatial attention (Experiment 2). Targets were light emitting diodes (LEDs) in the left and right hemispace of a visual display. Following left or right target illumination, participants reached to point-to or grasp target object in blocked trials. Coincident with target onset, a distractor LED illuminated in the same or opposite hemispace between the initiation point and target, or no distractor appeared. In Experiment 1, during grasping there was a larger temporal interference effect (slower reach initiation) than with pointing. When grasping versus pointing, participants deviated more towards same-side distractors and away from opposite-side distractors. In Experiment 2, distractors onset 200ms prior to (?200-ms), coincident with (0 ms), or 200ms following (+200 ms) the target. For both reach types, ?200-ms distractors had greater onset temporal interference than 0 ms and +200-ms distractors. For grasping, +200 ms distractors had larger temporal interference than 0 ms distractors. For ?200-ms, reach trajectories deviated more towards opposite-side distractors and away from same-side distractors, the reverse of the pattern for 0 ms and +200-ms distractors.     

Effect of spatial inhibition on saccade trajectory depends on location-based mechanisms

Abstract:  Saccade trajectory often curves away from a previously attended, inhibited location. A recent study of curved saccades showed that an inhibitory effect prevents ineffective reexamination during serial visual search. The time course of this effect differs from that of a similar inhibitory effect, known as inhibition of return (IOR). In the present study, we examined whether this saccade-related inhibitory effect can operate in an object-based manner (similar to IOR). Using a spatial cueing paradigm, we demonstrated that if a cue is presented on a placeholder that is then shifted from its original location, the saccade trajectory curves away from the original (cued) location (Experiment 1), yet the IOR effect is observed on the cued placeholder (Experiment 2). The inhibitory mechanism that causes curved saccades appears to operate in a location-based manner, whereas the mechanism underlying IOR appears to operate in an object-based manner. We propose that these inhibitory mechanisms work in a complementary fashion to guide eye movements efficiently under conditions of a dynamic visual environment.     

Developmental Differences in the Influence of Distractors on Maintenance in Spatial Working Memory

This study examined whether attention to a location plays a role in the maintenance of locations in spatial working memory in young children as it does in adults. This study was the first to investigate whether distractors presented during the delay of a spatial working-memory task influenced young children’s memory responses. Across 2 experiments, 3- and 6-year-olds completed a spatial working-memory task featuring a static target location and distractor location. Results indicated a change from 3 years to 6 years of age in how distractors influenced memory. Six-year-olds’ memory responses were biased away from a distractor that was close to the target location and on the outside of the target location relative to the center of the monitor. Distractors that were far from the target or that were toward the center of the monitor relative to the target location had no effect. Three-year-olds’ responses were biased toward a distractor when the distractor was on the outside of the target location and farther from the target. Distractors that were near the target location or toward the center of the monitor had no effect. These biases provide evidence that young children’s maintenance of a location in memory is influenced by attention.     

Presaccadic target competition attenuates distraction

Although it is well known that salient nontargets can capture attention despite being task irrelevant, several studies have reported short fixation dwell times, suggesting the presence of an attentional mechanism to “rapidly reject” dissimilar distractors. Rapid rejection has been hypothesized to depend on the strong mismatch between distractor features and the target template, but it is unknown whether the presence of strong feature mismatch is sufficient, or if the presence of a target at a competing location is also necessary. Here, we investigated this question by first replicating the finding of rapid rejection for dissimilar distractors in the presence of a concurrent target (Experiment 1); manipulating the onset of the target stimulus relative to the distractor (Experiment 2); and using a saccade-contingent display to delay the target onset until after the first saccade was initiated. The results demonstrate that the speed of distractor rejection depends on the presence of target competition prior to the initiation of the first saccade, and not after the saccade. This suggests that stimulus competition for covert attention sets a “saccade priority map” that unfolds over time, resulting in faster corrective saccades to an anticipated object with higher top-down attentional priority.     

Measuring saccade curvature: A curve-fitting approach

Saccade curvature is becoming a popular measure for detecting the presence of competing saccadic motor programs. Several different methods of quantifying saccade curvature have been employed. In the present study, we compared these metrics with each other and with novel measures based on curve fitting. Initial deviation metrics were only moderately associated with the more widely used metric of maximum curvature. The latter was strongly related to a recently developed area-based measure and to the novel methods based on second- and third-order polynomial fits. The curve-fitting methods showed that although most saccades curved in only one direction, there was a population of trajectories with both a maximum and a minimum (i.e., double-curved saccades). We argue that a curvature metric based on a quadratic polynomial fit deals effectively with both types of trajectories and, because it is based on all the samples of a saccade, is less susceptible to sampling noise.     

The adaptive character of the attentional system: statistical sensitivity in a target localization task

A localization task required participants to indicate which of 4 locations contained a briefly displayed target. Most displays also contained a distractor that was not equally probable in these locations, affecting performance dramatically. Responses were faster when a display had no distractor and almost as fast when the distractor was in its frequent location. Conversely, responses were slower when targets appeared in frequent-distractor locations, even thou targets were equally likely in each location. Negative-priming effects were reliably smaller when targets followed distractors in the frequent-distractor location compared to the rare-distractor location, challenging the episodic-retrieval account Experiment 2 added a 5th location that rarely displayed distractors and never targets, yet responses slowed most when distractors appeared there. The results confirmed that the attentional system is sensitive to first- and higher-order statistical patterns and can make short- and long-term adjustments in preferences based on prior history of inspecting unsuccessful locations.     

On varying the span of visual attention: Evidence for two modes of spatial attention

Three experiments investigated whether subjects could selectively attend to a target item presented in close spatial proximity to a distractor element. Typically, the display consisted of two curved lines, one a target and the other a distractor. The task was to judge the direction of curvature of the target. When subjects attended to a target away from fixation, performance was affected by the presence of a distractor within an area of 1° around the target. In contrast, the distractor did not influence target processing when subjects fixated on the target location. Two modes of visual attention are proposed. When a target is located away from fixation, a “wide” attentional span is employed. With the present stimuli, this led to the detailed processing of items within 1° of the attended position. However, when targets are at fixation a “narrow” span can be adopted, with the result that there is differential processing of attended items even within the formerly critical area.     

The effect of novel distractors on performance in focused attention tasks: A cognitive-psychophysiological approach

In the present study, we examined whether or not novel stimuli affect performance in a focused attention task. Participants responded to a central target while an irrelevant distractor in the visual display was occasionally changed. In Experiment 1, both target and distractor were presented centrally within the focus of attention. In Experiment 2, a central target was presented along with an irrelevant distractor at a peripheral location, outside the focus of attention. Novel distractors were associated with longer latencies and enhanced orienting responses (as measured by skin conductance responses) only when presented at an attended location. In contrast, as is demonstrated in Experiment 3, the same peripheral novel distractors interfered with task performance when they possessed task-relevant information. These results indicate that there is a fundamental difference between novel stimuli and task-relevant stimuli. Whereas the former exert influence only within the focus of attention, the latter affect performance even when positioned in an unattended location. Our findings have important implications for the operation of visual attention.     

Spatial working memory and inhibition of return

Recently we showed that maintaining a location in spatial working memory affects saccadic eye movement trajectories, in that the eyes deviate away from the remembered location (Theeuwes, Olivers, & Chizk, 2005). Such saccade deviations are assumed to be the result of inhibitory processes within the oculomotor system. The present study investigated whether this inhibition is related to the phenomenon of inhibition of return (IOR), the relatively slow selection of previously attended locations as compared with new locations. The results show that the size of IOR to a location was not affected by whether or not the location was kept in working memory, but the size of the saccade trajectory deviation was affected. We conclude that inhibiting working memory-related eye movement activity is not the same as inhibiting a previously attended location in space.     

What is learned in spatial contextual cuing--configuration or individual locations?

With the use of spatial contextual cuing, we tested whether subjects learned to associate target locations with overall configurations of distractors or with individual locations of distractors. In Experiment 1, subjects were trained on 36 visual search displays that contained 36 sets of distractor locations and 18 target locations. Each target location was paired with two sets of distractor locations on separate trials. After training, the subjects showed perfect transfer to recombined displays, which were created by combining half of one trained distractor set with half of another trained distractor set. This result suggests that individual distractor locations were sufficient to cue the target location. In Experiment 2, the subjects showed good transfer from trained displays to rescaled, displaced, and perceptually regrouped displays, suggesting that the relative locations among items were also learned. Thus, both individual target-distractor associations and configural associations are learned in contextual cuing.     

What is learned in spatial contextual cuing— configuration or individual locations?

With the use of spatial contextual cuing, we tested whether subjects learned to associate target locations with overall configurations of distractors or with individual locations of distractors. In Experiment 1, subjects were trained on 36 visual search displays that contained 36 sets of distractor locations and 18 target locations. Each target location was paired with two sets of distractor locations on separate trials. After training, the subjects showed perfect transfer to recombined displays, which were created by combining half of one trained distractor set with half of another trained distractor set. This result suggests that individual distractor locations were sufficient to cue the target location. In Experiment 2, the subjects showed good transfer from trained displays to rescaled, displaced, and perceptually regrouped displays, suggesting that the relative locations among items were also learned. Thus, both individual target-distractor associations and configural associations are learned in contextual cuing.     

A new look at novelty effects: Guiding search away from old distractors

We examined whether search is guided to novel distractors. In Experiment 1, subjects searched for a target among one new and a variable number of old distractors. Search displays in Experiment 2 consisted of an equal number of new, old, and familiar distractors (the latter repeated occasionally). We found that eye movements were preferentially directed to a new distractor on target-absent trials and that subjects tended to immediately fixate a new distractor after leaving the target on target-present trials. In both cases, first fixations on old distractors were consistently less frequent than could be explained by chance. We interpret these patterns as evidence for negative guidance: Subjects learn the visual features associated with the set of old distractors and then guide their search away from these features, ultimately resulting in the preferential fixation of novel distractors.