Axis Of Eye Velocity Research Articles

Coding of Velocity Storage in the Vestibular Nuclei.

Semicircular canal afferents sense angular acceleration and output angular velocity with a short time constant of ≈4.5 s. This output is prolonged by a central integrative network, velocity storage that lengthens the time constants of eye velocity. This mechanism utilizes canal, otolith, and visual (optokinetic) information to align the axis of eye velocity toward the spatial vertical when head orientation is off-vertical axis. Previous studies indicated that vestibular-only (VO) and vestibular-pause-saccade (VPS) neurons located in the medial and superior vestibular nucleus could code all aspects of velocity storage. A recently developed technique enabled prolonged recording while animals were rotated and received optokinetic stimulation about a spatial vertical axis while upright, side-down, prone, and supine. Firing rates of 33 VO and 8 VPS neurons were studied in alert cynomolgus monkeys. Majority VO neurons were closely correlated with the horizontal component of velocity storage in head coordinates, regardless of head orientation in space. Approximately, half of all tested neurons (46%) code horizontal component of velocity in head coordinates, while the other half (54%) changed their firing rates as the head was oriented relative to the spatial vertical, coding the horizontal component of eye velocity in spatial coordinates. Some VO neurons only coded the cross-coupled pitch or roll components that move the axis of eye rotation toward the spatial vertical. Sixty-five percent of these VO and VPS neurons were more sensitive to rotation in one direction (predominantly contralateral), providing directional orientation for the subset of VO neurons on either side of the brainstem. This indicates that the three-dimensional velocity storage integrator is composed of directional subsets of neurons that are likely to be the bases for the spatial characteristics of velocity storage. Most VPS neurons ceased firing during drowsiness, but the firing rates of VO neurons were unaffected by states of alertness and declined with the time constant of velocity storage. Thus, the VO neurons are the prime components of the mechanism of coding for velocity storage, whereas the VPS neurons are likely to provide the path from the vestibular to the oculomotor system for the VO neurons.

Three-dimensional kinematics of saccadic and pursuit eye movements in humans: Relationship between Donders’ and Listing’s laws

For Listing’s law to be obeyed during eye movements, the “half-angle rule” must be satisfied: the eye velocity axis must tilt away from Listing’s plane by half the angle of eye position eccentricity from primary position. We aimed to determine if this rule is satisfied during horizontal and vertical pursuit compared with saccades. Three-dimensional (3-d) eye rotation data were acquired from five normal head-fixed humans using the search coil technique. Saccades were recorded in response to 40° horizontal or vertical steps in target position, at different elevations and azimuths. Pursuit was recorded while tracking a target moving horizontally or vertically at 20°/s, with peak-to-peak amplitude of 40°, at the same elevations and azimuths. First- and second-order surfaces were fitted to 3-d eye position data from periods of fixation. In all subjects, eye positions did not lie on a planar surface, but on a twisted surface in 3-d space. The tilt-angle coefficient (TAC) during saccades and pursuit was calculated as the ratio of the angle of eye velocity axis tilt to the angle of eye position eccentricity. During horizontal saccades and pursuit, mean TACs were 0.58 and 0.64, respectively. During vertical saccades and pursuit, mean TACs were 0.35 and 0.43, respectively, and lower than their horizontal counterparts (p<0.05). These findings suggest that Listing’s law is not perfectly satisfied during saccades or pursuit. On the basis of model simulations, we propose that the discrepancy in horizontal and vertical TACs causes eye positions to lie on a twisted rather than a planar surface.

Three‐dimensional kinematics of saccadic eye movements in humans: is the “half‐angle rule” obeyed?

The "half-angle rule" must be satisfied for Listing's law to be obeyed during saccades: the eye velocity axis must tilt away from Listing's plane by half the angle of eye position eccentricity from primary position. We aimed to determine if this rule is satisfied during saccades. Horizontal and vertical saccades were recorded using the search coil technique, at different elevations and azimuths, in five normal humans. In all, eye positions were located on a twisted rather than planar surface. The mean tilt-angle coefficients (TACs; ratio of angle of eye velocity axis tilt to angle of eye position eccentricity) were 0.57 and 0.34 for horizontal and vertical saccades, respectively. TACs were significantly lower for vertical saccades. Thus, Listing's law is not perfectly obeyed during saccades. We suggest that the discrepancy in horizontal and vertical TACs causes eye positions to lie on a twisted rather than planar surface in three-dimensional space.

Kinematics of the Rotational Vestibuloocular Reflex: Role of the Cerebellum

We studied the effect of cerebellar lesions on the 3-D control of the rotational vestibuloocular reflex (RVOR) to abrupt yaw-axis head rotation. Using search coils, three-dimensional (3-D) eye movements were recorded from nine patients with cerebellar disease and seven normal subjects during brief chair rotations (200 degrees /s(2) to 40 degrees /s) and manual head impulses. We determined the amount of eye-position dependent torsion during yaw-axis rotation by calculating the torsional-horizontal eye-velocity axis for each of three vertical eye positions (0 degrees , +/-15 degrees ) and performing a linear regression to determine the relationship of the 3-D velocity axis to vertical eye position. The slope of this regression is the tilt angle slope. Overall, cerebellar patients showed a clear increase in the tilt angle slope for both chair rotations and head impulses. For chair rotations, the effect was not seen at the onset of head rotation when both patients and normal subjects had nearly head-fixed responses (no eye-position-dependent torsion). Over time, however, both groups showed an increasing tilt-angle slope but to a much greater degree in cerebellar patients. Two important conclusions emerge from these findings: the axis of eye rotation at the onset of head rotation is set to a value close to head-fixed (i.e., optimal for gaze stabilization during head rotation), independent of the cerebellum and once the head rotation is in progress, the cerebellum plays a crucial role in keeping the axis of eye rotation about halfway between head-fixed and that required for Listing's Law to be obeyed.

Rotational and translational optokinetic nystagmus have different kinematics

We studied the dependence of ocular torsion on eye position during horizontal optokinetic nystagmus (OKN) elicited by random-dot translational motion (tOKN) and prolonged rotation in the light (rOKN). For slow and quick phases, we fit the eye-velocity axis to vertical eye position to determine the tilt angle slope (TAS). The TAS for tOKN was 0.48 for both slow and quick phases, close to what is found during translational motion of the head. The TAS for rOKN was less for both slow (0.11) and quick phases (0.26), close to what is found during rotational motion of the head. Our findings are consistent with the notion that translational and rotational optic flow are processed differently by the brain and that they produce different 3-D eye movement commands that are comparable to the different commands generated in response to vestibular signals when the head is actually translating or rotating.

Eye-position dependence of torsional velocity during step-ramp pursuit and transient yaw rotation in humans

The time course of eye-position-dependent torsion during transient horizontal pursuit and yaw rotation was examined in seven normal human subjects. The stimuli consisted of step-ramp target motion (25, 40 degrees /s) and brief chair rotation (approximately 200 degrees /s(2) accelerated to 40 degrees /s) at three different vertical positions (center 0 degrees , up or down 15 degrees ). Three-dimensional eye movements were recorded with dual search coils. The kinematics of pursuit and the rotational vestibulo-ocular reflex (rVOR) were assessed by determining the tilt-angle slope, a measure of the variation of the axis of eye-velocity with vertical eye position. We found that the tilt-angle slope during pursuit was initially 0.4+/-0.07 (mean+/-95% confidence interval) and then gradually rose to 0.64+/-0.04, at about the time that the steady-state eye-velocity was reached. The rVOR began with a nearly head-fixed axis (0.08+/-0.04), appropriate for full retinal image stabilization, followed by a gradual increase of the tilt-angle slope to 0.31+/-0.02. Thus, differences between pursuit and the rVOR with respect to Listing's law can be seen from the onset of transient responses, although in both cases eye-position-dependent torsion increases with time. This temporal evolution of the axis of eye-velocity may involve the velocity-storage mechanism.

Cerebellar Disease Alters the Axis of the High-Acceleration Vestibuloocular Reflex

L. W. Schultheis and D. A. Robinson showed that the axis of the rotational vestibuloocular reflex (RVOR) cannot be altered by visual-vestibular mismatch ("cross-axis adaptation") when the vestibulocerebellum is lesioned. This suggests that the cerebellum may calibrate the axis of eye velocity of the RVOR under natural conditions. Thus we asked whether patients with cerebellar disease have alterations in the RVOR axis and, if so, what might be the mechanism. We used three-axis scleral coils to record head and eye movements during yaw, pitch, and roll head impulses in 18 patients with cerebellar disease and in a comparison group of eight subjects without neurologic disease. We found distinct shifts of the eye-velocity axis in patients. The characteristic finding was a disconjugate upward eye velocity during yaw. Measured at 70 ms after the onset of head rotation, the median upward gaze velocity was 15% of yaw head velocity for patients and <1% for normal subjects (P < 0.001). Upward eye velocity was greater in the contralateral (abducting) eye during yaw and in the ipsilateral eye during roll. Patients had a higher gain (eye speed/head speed) for downward than for upward pitch (median ratio of downward to upward gain: 1.3). In patients, upward gaze velocities during both yaw and roll correlated with the difference in anterior (AC) and posterior canal excitations, scaled by the respective pitch gains. Our findings support the hypothesis that upward eye velocity during yaw results from AC excitation, which must normally be suppressed by the intact cerebellum.

Ocular and perceptual responses to linear acceleration in microgravity: Alterations in otolith function on the COSMOS and Neurolab flights

In this paper we review space flight experiments performed by our laboratory. Rhesus monkeys were tested before and after 12 days in orbit on COSMOS flights 2044 (1989) and 2229 (1992-1993). There was a long-lasting decrease in post-flight ocular counter-rolling (70%) and vergence (50%) during off-vertical axis rotation. In one animal, the orientation of optokinetic after-nystagmus shifted by 28 degrees from the spatial vertical towards the body vertical early post-flight. Otolith-ocular and perceptual responses were also studied in four astronauts on the 17-day Neurolab shuttle mission (STS-90) in 1998. Ocular counter-rolling was unchanged in response to 1-g and 0.5-g Gy centrifugation during and after flight and to post-flight static roll tilts relative to pre-flight values. Orientation of the optokinetic nystagmus eye velocity axis to gravito-inertial acceleration (GIA) during centrifugation was also unaltered by exposure to microgravity. Perceptual orientation to the GIA was maintained in-flight, and subjects did not report sensation of translation during constant velocity centrifugation. These studies suggest that percepts and ocular responses to tilt are determined by sensing the body vertical relative to the GIA. The findings also raise the possibility that 'artificial gravity' during the Neurolab flight counteracted adaptation of these otolith-ocular responses.

Eye movements during multi-axis whole-body rotations.

The semi-circular canals and the otolith organs both contribute to gaze stabilization during head movement. We investigated how these sensory signals interact when they provide conflicting information about head orientation in space. Human subjects were reoriented 90 degrees in pitch or roll during long-duration, constant-velocity rotation about the earth-vertical axis while we measured three-dimensional eye movements. After the reorientation, the otoliths correctly indicated the static orientation of the subject with respect to gravity, while the semicircular canals provided a strong signal of rotation. This rotation signal from the canals could only be consistent with a static orientation with respect to gravity if the rotation-axis indicated by the canals was exactly parallel to gravity. This was not true, so a cue-conflict existed. These conflicting stimuli elicited motion sickness and a complex tumbling sensation. Strong horizontal, vertical, and/or torsional eye movements were also induced, allowing us to study the influence of the conflict between the otoliths and the canals on all three eye-movement components. We found a shortening of the horizontal and vertical time constants of the decay of nystagmus and a trend for an increase in peak velocity following reorientation. The dumping of the velocity storage occurred regardless of whether eye velocity along that axis was compensatory to the head rotation or not. We found a trend for the axis of eye velocity to reorient to make the head-velocity signal from the canals consistent with the head-orientation signal from the otoliths, but this reorientation was small and only observed when subjects were tilted to upright. Previous models of canal-otolith interaction could not fully account for our data, particularly the decreased time constant of the decay of nystagmus. We present a model with a mechanism that reduces the velocity-storage component in the presence of a strong cue-conflict. Our study, supported by other experiments, also indicates that static otolith signals exhibit considerably smaller effects on eye movements in humans than in monkeys.

Three-dimensional spatial characteristics of caloric nystagmus

We investigated the three-dimensional spatial characteristics of caloric nystagmus during excitation and inhibition of the lateral semicircular canal in five normal human subjects. Each subject was repositioned in 45 degrees steps at 1-min intervals such that the right lateral semicircular canal plane was reoriented in pitch, from 135 degrees backwards from the upright position to 135 degrees forwards, while the right ear was continuously stimulated with air at 44 degrees C. In orientations in which caloric stimulus resulted in excitation of the right lateral semicircular canal, the eye velocity axis was orthogonal to the average orientation of the right lateral semicircular canal plane. However, in orientations in which caloric stimulus resulted in inhibition of the right lateral semicircular canal, the eye velocity axis was orthogonal to the average orientation of the left and not the right lateral semicircular canal plane. These findings suggest that velocity and direction of caloric nystagmus depend not only on the absolute magnitude of vestibular activity on the stimulated side but also on the differences in activity between the left and right vestibular nuclei, most probably mediated centrally via brainstem commissural pathways.

Role of muscle pulleys in producing eye position-dependence in the angular vestibuloocular reflex: a model-based study.

It is well established that the head and eye velocity axes do not always align during compensatory vestibular slow phases. It has been shown that the eye velocity axis systematically tilts away from the head velocity axis in a manner that is dependent on eye-in-head position. The mechanisms responsible for producing these axis tilts are unclear. In this model-based study, we aimed to determine whether muscle pulleys could be involved in bringing about these phenomena. The model presented incorporates semicircular canals, central vestibular pathways, and an ocular motor plant with pulleys. The pulleys were modeled so that they brought about a rotation of the torque axes of the extraocular muscles that was a fraction of the angle of eye deviation from primary position. The degree to which the pulleys rotated the torque axes was altered by means of a pulley coefficient. Model input was head velocity and initial eye position data from passive and active yaw head impulses with fixation at 0 degrees, 20 degrees up and 20 degrees down, obtained from a previous experiment. The optimal pulley coefficient required to fit the data was determined by calculating the mean square error between data and model predictions of torsional eye velocity. For active head impulses, the optimal pulley coefficient varied considerably between subjects. The median optimal pulley coefficient was found to be 0.5, the pulley coefficient required for producing saccades that perfectly obey Listing's law when using a two-dimensional saccadic pulse signal. The model predicted the direction of the axis tilts observed in response to passive head impulses from 50 ms after onset. During passive head impulses, the median optimal pulley coefficient was found to be 0.21, when roll gain was fixed at 0.7. The model did not accurately predict the alignment of the eye and head velocity axes that was observed early in the response to passive head impulses. We found that this alignment could be well predicted if the roll gain of the angular vestibuloocular reflex was modified during the initial period of the response, while pulley coefficient was maintained at 0.5. Hence a roll gain modification allows stabilization of the retinal image without requiring a change in the pulley effect. Our results therefore indicate that the eye position-dependent velocity axis tilts could arise due to the effects of the pulleys and that a roll gain modification in the central vestibular structures may be responsible for countering the pulley effect.

Vertical eye position-dependence of the human vestibuloocular reflex during passive and active yaw head rotations.

Vertical eye position-dependence of the human vestibuloocular reflex during passive and active yaw head rotations. The effect of vertical eye-in-head position on the compensatory eye rotation response to passive and active high acceleration yaw head rotations was examined in eight normal human subjects. The stimuli consisted of brief, low amplitude (15-25 degrees ), high acceleration (4,000-6,000 degrees /s2) yaw head rotations with respect to the trunk (peak velocity was 150-350 degrees /s). Eye and head rotations were recorded in three-dimensional space using the magnetic search coil technique. The input-output kinematics of the three-dimensional vestibuloocular reflex (VOR) were assessed by finding the difference between the inverted eye velocity vector and the head velocity vector (both referenced to a head-fixed coordinate system) as a time series. During passive head impulses, the head and eye velocity axes aligned well with each other for the first 47 ms after the onset of the stimulus, regardless of vertical eye-in-head position. After the initial 47-ms period, the degree of alignment of the eye and head velocity axes was modulated by vertical eye-in-head position. When fixation was on a target 20 degrees up, the eye and head velocity axes remained well aligned with each other. However, when fixation was on targets at 0 and 20 degrees down, the eye velocity axis tilted forward relative to the head velocity axis. During active head impulses, the axis tilt became apparent within 5 ms of the onset of the stimulus. When fixation was on a target at 0 degrees, the velocity axes remained well aligned with each other. When fixation was on a target 20 degrees up, the eye velocity axis tilted backward, when fixation was on a target 20 degrees down, the eye velocity axis tilted forward. The findings show that the VOR compensates very well for head motion in the early part of the response to unpredictable high acceleration stimuli-the eye position- dependence of the VOR does not become apparent until 47 ms after the onset of the stimulus. In contrast, the response to active high acceleration stimuli shows eye position-dependence from within 5 ms of the onset of the stimulus. A model using a VOR-Listing's law compromise strategy did not accurately predict the patterns observed in the data, raising questions about how the eye position-dependence of the VOR is generated. We suggest, in view of recent findings, that the phenomenon could arise due to the effects of fibromuscular pulleys on the functional pulling directions of the rectus muscles.

Effects of tilt of the gravito-inertial acceleration vector on the angular vestibuloocular reflex during centrifugation.

Effects of tilt of the gravito-inertial acceleration vector on the angular vestibuloocular reflex during centrifugation. Interaction of the horizontal linear and angular vestibuloocular reflexes (lVOR and aVOR) was studied in rhesus and cynomolgus monkeys during centered rotation and off-center rotation at a constant velocity (centrifugation). During centered rotation, the eye velocity vector was aligned with the axis of rotation, which was coincident with the direction of gravity. Facing and back to motion centrifugation tilted the resultant of gravity and linear acceleration, gravito-inertial acceleration (GIA), inducing cross-coupled vertical components of eye velocity. These components were upward when facing motion and downward when back to motion and caused the axis of eye velocity to reorient from alignment with the body yaw axis toward the tilted GIA. A major finding was that horizontal time constants were asymmetric in each monkey, generally being longer when associated with downward than upward cross coupling. Because of these asymmetries, accurate estimates of the contribution of the horizontal lVOR could not be obtained by simply subtracting horizontal eye velocity profiles during facing and back to motion centrifugation. Instead, it was necessary to consider the effects of GIA tilts on velocity storage before attempting to estimate the horizontal lVOR. In each monkey, the horizontal time constant of optokinetic after-nystagmus (OKAN) was reduced as a function of increasing head tilt with respect to gravity. When variations in horizontal time constant as a function of GIA tilt were included in the aVOR model, the rising and falling phases of horizontal eye velocity during facing and back to motion centrifugation were closely predicted, and the estimated contribution of the compensatory lVOR was negligible. Beating fields of horizontal eye position were unaffected by the presence or magnitude of linear acceleration during centrifugation. These conclusions were evaluated in animals in which the low-frequency aVOR was abolished by canal plugging, isolating the contribution of the lVOR. Postoperatively, the animals had normal ocular counterrolling and horizontal eye velocity modulation during off-vertical axis rotation (OVAR), suggesting that the otoliths were intact. No measurable horizontal eye velocity was elicited by centrifugation with angular accelerations </=40 degrees /s2 and angular velocities </=400 degrees /s. We conclude that in rhesus and cynomolgus monkeys, differences between horizontal eye velocities recorded during facing and back to motion constant velocity centrifugation can be explained by orienting effects of the GIA tilt on the time constants of the horizontal aVOR and not by a superposed lVOR.

Off-center yaw rotation: effect of naso-occipital linear acceleration on the nystagmus response of normal human subjects and patients after unilateral vestibular loss.

Dual search coils were used to record horizontal, vertical and torsional eye movement components of one eye during nystagmus caused by off-center yaw rotation (yaw centrifugation). Both normal healthy human subjects (n=7) and patients with only one functioning labyrinth (n=12) were studied in order to clarify how the concomitant linear acceleration affected the nystagmus response. Each subject was seated with head erect on the arm of a fixed-chair human centrifuge, 1 m away from the center of the rotation, and positioned to be facing along a radius; either towards (facing-in) or away from (facing-out) the center of rotation. Both yaw right and yaw left angular accelerations of 10 degrees s(-2) from 0 to 200 degrees/s were studied. During rotation a centripetal linear acceleration (increasing from 0 to 1.24xg units) was directed along the subject's naso-occipital axis resulting in a shift of the resultant angle of the gravitoinertial acceleration (GIA) of 51 degrees in the subject's pitch plane and an increase in the total GIA magnitude from 1.0 to 1.59xg. In normal subjects during the angular acceleration off-center there were, in addition to the horizontal eye velocity components, torsional and vertical eye velocities present. The magnitude of these additional components, although small, was larger than observed during similar experiments with on-center angular acceleration (Haslwanter et al. 1996), and the change in these components is attributed to the additional effect of the linear acceleration stimulation. In the pitch plane the average size of the shift of the axis of eye velocity (AEV) during the acceleration was about 8 degrees for a 51 degrees shift of the GIA (around 16% of the GIA shift) so that the AEV-GIA alignment was inadequate. There was a very marked difference in the size of the AEV shift depending on whether the person was facing-in [AEV shift forward (i.e. non-compensatory) of about 4 degrees] or facing-out [AEV shift forward (i.e. compensatory) of around 12 degrees]. The linear acceleration decreased the time constant of decay of the horizontal component of the post-rotatory nystagmus: from an average of 24.8 degrees/s facing-in to an average of 11.3 degrees/s facing-out. The linear acceleration dumps torsional eye velocity in an manner analogous to, but independent of, the dumping of horizontal eye velocity. Patients with UVD had dramatically reduced torsional eye velocities for both facing-in and facing-out headings, and there was little if any shift of the AEV in UVD patients. The relatively small effects of linear acceleration on human canal-induced nystagmus found here confirms other recent studies in humans (Fetter et al. 1996) in contrast to evidence from monkeys and emphasizes the large and important differences between humans and monkeys in otolith-canal interaction. Our results confirm the vestibular control of the axis of eye velocity of humans is essentially head-referenced whereas in monkeys that control is essentially space-referenced.

Contribution of vestibular commissural pathways to spatial orientation of the angular vestibuloocular reflex.

During nystagmus induced by the angular vestibuloocular reflex (aVOR), the axis of eye velocity tends to align with the direction of gravito-inertial acceleration (GIA), a process we term "spatial orientation of the aVOR." We studied spatial orientation of the aVOR in rhesus and cynomolgus monkeys before and after midline section of the rostral medulla abolished all oculomotor functions related to velocity storage, leaving the direct optokinetic and vestibular pathways intact. Optokinetic afternystagmus and the bias component of off-vertical-axis rotation were lost, and the aVOR time constant was reduced to a value commensurate with the time constants of primary semicircular canal afferents. Spatial orientation of the aVOR, induced either during optokinetic or vestibular stimulation, was also lost. Vertical and roll aVOR time constants could no longer be lengthened in side-down or supine/prone positions, and static and dynamic tilts of the GIA no longer produced cross-coupling from the yaw to pitch and yaw to roll axes. Consequently, the induced nystagmus remained entirely in head coordinates after the lesion, regardless of the direction of the resultant GIA vector. Gains of the aVOR and of optokinetic nystagmus to steps of velocity were unaffected or slightly increased. These results are consistent with a model in which the direct aVOR pathways are organized in semicircular canal coordinates and spatial orientation is restricted to the indirect (velocity storage) pathways.

The three-dimensional human vestibulo-ocular reflex: response to long-duration yaw angular accelerations.

We recorded three-dimensional eye movements during angular acceleration steps from 0 to 250 degrees/s at 20 degrees/s2 about an earth-vertical axis. Experiments were performed on 27 normal subjects and on 19 patients who had recovered well from unilateral vestibular deafferentation on the right or left side. In addition to compensatory horizontal eye movements, significant vertical and torsional eye movement components were elicited. These vertical and torsional eye velocity traces led to a shift of the axis of eye velocity away from the axis of head velocity. Horizontal, vertical, and torsional velocity components showed clear differences between normals and patients with unilateral vestibular deafferentation. In normals, the axis of eye velocity tilted backward and slightly away from the axis of head velocity. Patients showed similar, but more pronounced, shifts during rotations toward the intact ear and shifts in the opposite direction for rotations toward the operated ear. Eye velocity traces were analyzed with special consideration given to the orientation of the axis of eye velocity. We speculate that the vertical and torsional velocity components may be due to the effects of Listing's plane, as well as the contributions of the otolith signals.

Torsional eye velocity components during yaw angular acceleration identify the side of unilateral vestibular deafferentation.

Using dual torsion scleral search coils we have recorded 3-dimensional eye position during yaw angular accelerations of 20 degrees/s2 about an earth vertical axis in healthy subjects and in patients with unilateral vestibular deafferentation (UVD). These experiments produced two interesting results: i) even in healthy subjects, the axis of eye velocity did not coincide with the (earth vertical) stimulus axis during centred rotation; ii) Patients with UVD had torsional eye velocity components that were systematically different from those in normal subjects. While in normals the direction of the torsional component of the eye velocity depended on the direction of rotation and was on average approximately symmetric for CW and CCW yaw rotation, there was a clear asymmetry in patients, which was distinctly different for left and right UVD.