Saccadic Control System Research Articles

A Stochastic Optimal Control Model with Internal Feedback and Velocity Tracking for Saccadic Eye Movements

A new stochastic optimal control model with internal feedback and velocity tracking is presented in this paper to study saccadic eye movement control. Recent evidence from neurophysiological studies of superior colliculus suggests the presence of a dynamic input to the saccade generation system that encodes saccade velocity, rather than just the static information of the desired saccade amplitude and direction. The new evidence makes it imperative to test if the saccadic control system can use a desired velocity input to achieve accurate behavioral outcomes. To test such a velocity-based architecture of saccade control, a new optimal control tracking model that incorporated two unique characteristics of internal feedback and stochasticity has been developed. The proposed model was validated using behavioral data of saccades generated by healthy human subjects in an experimental setup. The model was able to generate displacement and velocity trajectories of horizontal saccades made to different amplitudes and predict saccades made to vertical and oblique directions. It captured the main sequence relationship between saccade amplitude and peak velocities that were observed in the behavioral data. This paper presents a novel optimal control framework with tracking and internal feedback and proposes the first-ever model of the saccadic system that uses an alternate interpretation of velocity-based control, contrary to the dominant end-point based models available in the literature.

Saccadic Eye Movements Are Associated With a Family History of Alcoholism at Baseline and After Exposure to Alcohol

Objective To evaluate the influence of family history of alcoholism (FHA) on the response of saccadic eye movements to alcohol.Method Saccadic performance was evaluated in 54 healthy adult subjects with a FHA (family history–positive) and 49 controls (family history–negative). Alcohol and placebo sessions were presented in counterbalanced order. Alcohol was administered intravenously to achieve and maintain a target breath alcohol concentration of 60 mg/100 ml (60%) for 160 min in each subject. During each session, saccadic eye movement testing was performed at baseline (before infusion of alcohol) and twice during the steady‐state target breath alcohol concentration. The saccadic testing elicited visually guided saccades (VGS) and antisaccades (AS). Saccadic latency and velocity and the percentage of AS errors were quantified and analyzed using multivariate analysis of variance.Results The family history–positive and family history–negative groups showed an overall difference at baseline in AS and VGS latencies and velocities in the alcohol and placebo sessions (p= 0.006). Alcohol delayed saccades such that AS and VGS latencies increased (p= 0.0001) and slowed the execution of saccades such that peak velocities decreased (p= 0.0002). The percentage of AS errors decreased after alcohol administration, but no significant effect of alcohol (alcohol versus placebo session) was observed (p= 0.1). Latency of AS saccades demonstrated a significant overall FHA effect (p= 0.02) and a significant interaction between FHA and response to alcohol over time (p= 0.02).Conclusions Differences in operational characteristics of the saccadic control system are associated with FHA in adult social drinkers, both at baseline and when the brain is exposed to ethanol at 60 mg/100 ml.

A distributed model of the saccadic system: The effects of internal noise

A neural network model for the saccadic control system was proposed recently. In this model, the superior colliculus (SC) was represented as a two-layered neural network, with the second (motor) layer having extensive lateral interconnections. The SC network then provided a distributed dynamic control signal to a lumped model of the brainstem burst generator. In this paper, the saccadic model is modified so that it more closely reproduces the behavior measured experimentally in the primate saccadic system. The burst generator in the earlier model was replaced by a modified version that bears a stronger resemblance to the primate burst generator. The artificial trigger signal of the earlier work was replaced by a more neurophysiologically plausible mechanism, in which temporal initiation of saccadic eye movements is achieved through the output of the SC network itself. With the help of a new training algorithm that simultaneously updated all feedforward and feedback connection strengths, the revised model was trained not only to elicit realistic horizontal and oblique simulated saccades, but also to produce more realistic activity in the model's motor layer units. Finally, temporal noise was incorporated into our model and further changes were made so that discharges of the motor layer units had the same amount of variability as that recorded in neural discharges in the primate SC. The performance of the model in the presence of the injected noise was analyzed for different saccadic paradigms. In each case, the degree of scatter in the simulated eye movements resembled that recorded in monkey under similar behavioral conditions. Based on our results, we draw some potentially important inferences about the operation of the actual saccadic eye movement control system.

Chasing and pursuit in the dolichopodid fly Poecilobothrus nobilitatus

1. Male Poecilobothrus nobilitatus show two distinct kinds of pursuit. Females are “shadowed” at a distance of a few cm, using both rotational and lateral movements. Other males are chased, in a pursuit that involves only rotation and fast forward flight. The rotational component of pursuit appears to have the same control system in both types of tracking, and it is best described as a continuous translation of the error angle between the direction of the target and the pursuing fly's body axis into the pursuing fly's angular velocity. The constant of proportionality is 30–40°·s−1 per degree, and the delay in the system is about 15 ms. Pursuit on the ground is 2–3 times slower than in flight, although the delay seems to be similar. 2. Attempts were made to see whether the aerial pursuits could be modelled effectively by a saccadic or discontinuous control system, as suggested for Musca pursuit (Wagner 1986). It was found that the velocity profiles of the chases could be fitted by an overlapping series of plausible saccade-like events, However, the correlation between visual information (error angle and error angular velocity) available just before each fictive saccade correlated poorly with saccade peak velocity. It is thus concluded that Poedlobothrus pursuit is basically continuous in nature, but it is argued that the two types of mechanism are hard to distinguish in natural behaviour.

Adaptation of aimed arm movements to sensorimotor discordance: evidence for direction-independent gain control

Human subjects pointed, without sight of their arm, at visual targets presented on a mirror-viewed monitor screen. During the adaptation period of each experiment, the position of the pointing fingertip was continuously recorded and displayed on the screen along with the targets. This visual feedback was not always veridical; rather, it was manipulated to require a gradual modification of the pointing response gain throughout the adaptation period. No visual feedback at all was available during the pre- and postadaptation periods of each experiment. The adaptive effect was determined as difference between pre- and postadaptation gains. In Expt. A, visual feedback during the adaptation period prescribed a gradual reduction of the horizontal response gain without specifying the gain for other directions; the adaptive effect was found to generalize uniformly to all movement directions. Expt. B1 prescribed a reduction of the horizontal, and an unchanged vertical gain component: in spite of this differential requirement, the adaptive effect was again uniform for all directions. Expt. B2 prescribed a reduction of the horizontal, and an increase of the vertical gain component: we found a reduced gain for all directions, with a mild direction-dependence in the magnitude of the adaptive effect. In a modified version of Expt. B2, no intermanual transfer of the adaptive effect was found. Expt. C1-3 prescribed gain reduction for target directions within 15, 30, or 45 degrees around the horizontal, and gain increase for all other directions: we found little or no adaptive effects under such conditions. From the above findings, we concluded that the adapted system controls movement gain largely independent of movement direction. This mechanism responds readily to requirements for gain reduction, but not gain increase. No evidence for an organization of the arm motor system in direction-selective channels was found, in contrast to findings on the saccadic control system in a paradigm similar to our Expt. A8. This discrepancy supports the view that arm and eye movements are controlled by distinct mechanisms.

ENDURING DYSMETRIA AND IMPAIRED GAIN ADAPTIVITY OF SACCADIC EYE MOVEMENTS IN WALLENBERG'S LATERAL MEDULLARY SYNDROME

Saccadic eye movements and the adaptive control of their amplitudes were examined in patients with Wallenberg's lateral medullary syndrome. Half of the patients had permanent saccadic dysmetria. Their primary saccades had asymmetric amplitudes: those made in response to an ipsilateral target step (i.e. to the lesion side) tended to be hypermetric and saccades made in response to a contralateral target step were strongly hypometric. Multiple correction saccades were needed for target fixation. The adjustment of the amplitude of artificially induced hypermetric saccades, called gain adaptivity, was examined experimentally by using double target steps. The first target step elicited the primary saccade which triggered a further target displacement. This second, intra-saccadic target displacement was opposite to the first target step and caused the primary saccade to overshoot the final target position. In this way a post-saccadic target position error was generated which had to be corrected for foveal fixation. With repetition of this stimulus sequence the saccadic control system of normal subjects made an adjustment in amplitude of the main saccade such that the overshooting gradually diminished. After a few hundred trials primary saccades became orthometric with respect to the final target position; in respect to the first target step they were, however, strongly hypometric. The experimental data show that patients with Wallenberg's syndrome had a reduced capability to readjust saccadic amplitude. This observation together with the enduring saccadic dysmetria suggest that adaptive gain control of saccades is impaired in patients with lesions restricted to the dorsolateral medulla. It is speculated that these lesions most likely disrupt olivo-cerebellar pathways which are believed to be of paramount importance in visuo-motor adaptation of the cerebellum.

Saccade control in a simulated robot camera-head system: neural net architectures for efficient learning of inverse kinematics

The high speed of saccades means that they cannot be guided by visual feedback, so that any saccadic control system must know in advance the correct output signals to fixate a particular retinal position. To investigate neural-net architectures for learning this inverse-kinematics problem we simulated a 4 deg-of-freedom robot camera-head system, in which the head could pan and tilt and the cameras pan and verge. The main findings were: (1) Linear nets, multilayer perceptrons (MLPs) trained by backpropagation, and cerebellar model arithmetic computers (CMACs) all learnt rapidly to 5-10% accuracy when given perfect error feedback. (2) For additional accuracy (down to 2%) two-layer nets learnt much faster than a single MLP or CMAC: the best combination tried was to have a CMAC learn the errors of a trained linear net. (3) Imperfect error signals were provided by a crude controller whose output was simply proportional to retinal input in the relevant axis, thereby providing a mechanism for (a) controlling the camera-head system when the feedforward neural net controller was wrong or inoperative, and (b) converting sensory error signals into motor error signals as required in supervised learning. It proved possible to train neural-net controllers using these imperfect error signals over a range of learning rates and crude-controller gains. These results suggest that appropriate neural-net architectures can provide practical, accurate and robust adaptive control for saccadic movements. In addition, the arrangement of a crude controller teaching a sophisticated one may be similar to that used by the primate saccadic system, with brainstem circuitry teaching the cerebellum.

Topography of eye-position sensitivity of saccades evoked electrically from the cat's superior colliculus

Saccades evoked electrically from the deep layers of the superior colliculus have been examined in the alert cat with its head fixed. Amplitudes of the vertical and horizontal components varied linearly with the starting position of the eye. The slopes of the linear-regression lines provided an estimate of the sensitivity of these components to initial eye position. In observations on 29 sites in nine cats, the vertical and horizontal components of saccades evoked from a given site were rarely influenced to the same degree by initial eye position. For most sites, the horizontal component was more sensitive than the vertical component. Sensitivities of vertical and horizontal components were lowest near the representations of the horizontal and vertical meridians, respectively, of the collicular retinotopic map, but otherwise exhibited no systematic retinotopic dependence. Estimates of component amplitudes for saccades evoked from the center of the oculomotor range also diverged significantly from those predicted from the retinotopic map. The results of this and previous studies indicate that electrical stimulation of the cat's superior colliculus cannot yield a unique oculomotor map or one that is in register everywhere with the sensory retinotopic map. Several features of these observations suggest that electrical stimulation of the colliculus produces faulty activation of a saccadic control system that computes target position with respect to the head and that small and large saccades are controlled differently.

Chasing with a Model Eye

ABSTRACT A simulation model for Polyphemus eye movements was based on an openloop saccadic control system, derived from previous observations of eye movements in response to sinusoidally oscillating targets. The model was checked against new observations of Polyphemus eye movements in response to erratically moving targets. It was used to predict eye movements in response to stimulus patterns deduced from an analysis of video-recordings of Polyphemus chases. The results show that the system is limited to tracking slow target movements, but that this is consistent with observed chasing behaviour.

異常眼球運動―ｏｐｓｏｃｌｏｎｕｓとｆｌｕｔｔｅｒ‐ｌｉｋｅ ｏｓｃｉｌｌａｔｉｏｎ―の１症例

A 67-year-old woman noticed tremor of her hand and jerking of her eyeballs on May 20, 1986, and then became unable to stand. The cerebrospinal changes and Magnetic Response Imaging (MRI) findings were compatible with encephalitis. She was treated with decadron, and the tremor and ocular oscillations gradually decreased.The abnormal ocular movements which were recorded by electro-oculography (EOG) on June 24, revealed frequent, irregular, rapid, multidirectional conjugate ocular oscillations (so-called opsoclonus) with the eye open or closed. These ocular oscillations were suppressed with Frenzel's spectacles. Gradually, the vertical components of the opsoclonus decreased. On June 26, in addition to the opsoclonus, pendullar oscillation consisting of approximately equal amplitude (so-called flutter-like oscillations) developed. The small ocular oscillations similar to flutter like oscillation were recorded by EOG and a fine intension tremor was still present after recovery. The latency of the saccade was prolonged by interruption of the abnormal ocular movements.It seems that the opsoclonus and flutter-like oscillation in this patient represented tremor of the eyeballs caused by a disorder of the burst cells in the brain stem saccadic control system in addition to a disturbance of the cerebellar dentate nuclear system.

Effects of eye position on saccades evoked electrically from superior colliculus of alert cats.

Electrical stimulation was carried out in the intermediate and deep gray layers of the superior colliculus in alert cats. The heads of the animals were fixed, and their eye movements were recorded with the scleral search coil method. Stimulation in the anterior two-thirds of the colliculus with long-duration pulse trains produced multiple saccades, as in the primate (45, 51), but their directions and amplitudes were influenced significantly by the initial position of the eye. Stimulation in the posterior part of the colliculus evoked saccades that appeared to be "goal-directed," whereas stimulation at the extreme caudal edge of the colliculus yielded centering saccades. These observations confirm previous reports of Roucoux and Crommelinck (48) and Guitton et al. (24). Saccades evoked during bilateral simultaneous stimulation of the superior colliculi were also dependent on the initial position of the eye. At certain relative intensities of stimulation on the two sides, saccades failed to occur when the eye was within a particular part of the oculomotor range. When the eye was outside this region, the same stimuli triggered an eye movement that drove the eye toward the zone of saccade failure. These findings indicate that saccadic commands resulting from focal collicular stimulation in the cat can be modified by information about current eye position. It is not certain where in the brain this occurs or by what neural mechanisms, but a local feedback model of the saccadic control system (46) can account for the main observations. The functional significance of these findings depends in large measure on the degree to which focal collicular stimulation reproduces naturally occurring patterns of neural activity.

Saccade abnormalities in patients with ocular flutter.

In a group of 8 patients with opsoclonus or ocular flutter, the eye movements were recorded by electro-oculography (EOG). The spontaneous eye movement pattern and the amplitude and peak velocity of the refixation saccade were analysed. The EOG recording demonstrated frequent bursts of horizontally and vertically directed saccades elicited without any intersaccadic interval. Three patients also demonstrated an increased frequency of square waves. In 6 of the 8 patients the peak velocity of the voluntary saccades was increased; in 5 patients this condition was found for saccades in both directions. Over-shooting oscillations or slightly hypermetric voluntary saccades occurred in 5 patients. It is proposed that the ocular flutter and the increased saccadic velocities found in the present series of patients is caused by a disorder of the burst cells in the brain stem saccadic control system.

Ｓｑｕａｒｅ ｗａｖｅ ｊｅｒｋｓの発現機序に関する臨床的検討

Unusual and bizarre abnormalities of ocular movement are often valuable localizing signs in central nervous system diseases. Square wave jerks (SWJ) are sporadic horizontal conjugate saccades away from the intended position of fixation, followed some 200 msec later by saccadic return to the fixation. SWJ are recognized as a normal phenomenon when present only under closed eyelids. SWJ during fixation are considered to be a pathological sign, indicative of cerebellar disease. Reports of SWJ in progressive supranuclear palsy, multiple sclerosis, posttraumatic cerebral syndrome, presenile dementia, and cerebral tumors suggest that brain stem or cerebral hemispheric disease may also be responsible. The mechanism of SWJ, however, remains unceretain. The present paper reports six cases of SWJ in two patients with cerebellar infarction, one with cerebellar hemorrhage, two with spinocerebellar degeneration, and one with dentato-pallidal degeneration. The localizing value of SWJ and its mechanism of action are discussed.SWJ were conjugate, horizontal, symmetrical, occurred in burtst of several seconds, had amplitudes of 4° to 35°, and were evoked whenever the patient attempted to shift visual fixation or pursue a moving target. Electronystagmographic recordings in one patient with cerebellar hemorrhage showed the following features of this disorder of saccadic eye movement : 1) SWJ was composed of saccades, 2) their frequency was 0.5 to 2 Hz, 3) bursts occurred with decreasing amplitude and 4) the eye position did not exhibit systematic drift during the intersaccadic period. These features reflect the increased gain and instability of the visually guided saccadic system. Our resuls suggest that the loss of the calibrator function of the cerebellum accounts for the development of the abnormality underlying SWJ, and that SWJ are the result of unwanted supranuclear trigger signals that interrupt the saccadic pulse-step control system thereby releasing saccadic burst units.

An adaptive control model for human head and eye movements while walking

When a person walks the head undergoes horizontal and vertical rotations, and also horizontal and vertical translations. To visually fixate on an object while walking, compensatory horizontal and vertical eye movements must be made. The model for this gaze (head plus eye) control system includes three types of eye movements: smooth pursuit, saccadic, and vestibulo-ocular. The smooth pursuit system uses a target-selective adaptive controller that compensates for the large inherent time delay and produces zero-latency tracking of predictable targets. Target movements were selected to minimize the role of the saccadic control system. Typical tracking is shown while seated with the head restrained, standing with unrestrained head, performing voluntary head rotations, and walking. Each additional degree of freedom produced additional head movements that were compensated by additional eye movements. It is shown that while a human walks the effects of head rotations (yaw) cancel the effects of head translations, thus minimizing the resulting horizontal eye rotations necessary to maintain fixation.

Disorders in cerebellar ocular motor control. I. Saccadic overshoot dysmetra, an oculographic, control system and clinico-anatomic analysis. II. Macrosaccadic oscillation, an oculographic, control system and clinico-anatomic analysis: by J.B. Selhorst, L. Stark, A.L. Ochs, and W.F. Hoyt. Brain 99:497–508, 509–522, 1976

Disorders in cerebellar ocular motor control. I. Saccadic overshoot dysmetra, an oculographic, control system and clinico-anatomic analysis. II. Macrosaccadic oscillation, an oculographic, control system and clinico-anatomic analysis: by J.B. Selhorst, L. Stark, A.L. Ochs, and W.F. Hoyt. Brain 99:497–508, 509–522, 1976

The accuracy of two-dimensional saccades in the absence of continuing retinal stimulation

iReceived 2 February 1977; in re~ised~~~ IS .&far& 1977) The ability of observers to execute saccades in the dark to previously visible targets has been demonstrated by Becker and Fuchs (1969) and Becker and Klein (1973). Becker and Fuchs report the occurrence of horizontal saccades when a target, 40’ away from fixation. was extinguished 350 msec prior to the initiation of the saccade. Becker and Klein report that observers, after practicing horizontal saccades to targets were able to continue making goal-directed sac&es in total darkness. The accuracy of these saccades, however, decreased with time. These results imp1y that continuing retinal stimulation is not necessary for the execution of goal-directed saccades. Rather, the saccadic control system appears able to utilize some sort of stored information. The present experiment assessed the accuracy with which the saccadic control system could execute, from memory alone, a consecutive series of two-dimensional saccades to targets randomly located in the fronto-parallel plane.

Some observations concerning saccadic eye movements

Eye movements, made in response to a target that was displaced twice (pulse-step stimuli), were recorded for three observers in order to ascertain the characteristics of the saccadic control system. Five types of pulse-step stimuli were randomly presented to the observers. The results indicate that observers did not always respond to the initial displacement of the target, but sometimes responded only to the target's final position. The percentage of instances in which observers responded to the pulse increased as the duration of the pulse increased from 50 to 200 msec. In those instances when observers responded to both target displacements, the latency of the second saccade was shorter than that of the first. When observera only responded to the target's final position, the latency of the saccade was shorter if the target's final position was in the same direction than if it was in the opposite direction from its initial displacement. These results suggest that there is continuous input of information into the saccadic control system.