VISUAL INFLUENCE ON HEAD SHAKING USING THE VESTIBULAR ... · The Vestibular Autorotation Test (V AT) is a computerized high-frequency test that uses a se ries of audible tones sweeping

Journal of Vestibular Research, Vol. 6, No.6, pp. 411-422, 1996 Published 1996 by Elsevier Science lnc.

Printed in the USA. All rights reserved 0957-4271196 $15.00 + .00

ELSEVIER PH S0957-4271(96)00072-9

Original Contribution

VISUAL INFLUENCE ON HEAD SHAKING USING THE VESTIBULAR AUTOROTATION TEST

Bob Cheung, Ken Money, and Paul Sarkar

Defence and Civil Institute of Environmental Medicine, North York, Ontario, Canada Reprint address: Bob Cheung, Ph.D., Spatial Disorientation, Aerospace Life Support Sector,

Defence and Civil Institute of Environmental Medicine, 1133 Sheppard Ave. W., P.O. Box 2000, North York, Ontario, Canada, M3M 3B9. Tel: (416)635-2053; Fax: (416)635-2204

o Abstract - In this study we investigated the vestibular system by recording eye movements in response to voluntary high-frequency head-only movements using the Vestibular Autorotation Test (V AT; Western System Research Inc., Los Angeles, California). Our objective was to evalu­ate if the VAT could be implemented as one of the screening tests for vestibular integrity in ~ircrews and potential pilots. We attempted to record hori­zontal and vertical eye movements using electro­oculography and head velocity with calibrated ro­tational velocity sensors. The gain and phase of the input and output signals were computed by discrete Fourier analysis. Seated subjects were in­structed to fixate on a real or imaginary target while making smooth head oscillations about the spinal axis in time to an audible cue from 0.5 to 6.0 Hz during an IS-second test period. Test trials in­cluded two conditions in the light with subjects fixated on a real target (el) or on an imaginary target on a blank screen ee2); three conditions in the dark in which subjects fixated on an imagi­nary target (C3), fixated on a remembered LED target in the dark after it was extinguished or fixated on a real target All thE riad~ t~'iah

were performed after dark adaptation for 30 min­utes. We were not able to obtain consistent verti~ cal VOR response (when the subjects oscillated their head about the inter aural axis) using the VAT. For horizontal eye movements from 2.0 to just over 4.7 Hz, when subjects fixated on an imaginary target, there was an unexpected and

The findings of this study were presented at the 66th Annual Scientific Meeting of the Aerospace Medical Association, May 7-11, 1995, Anaheim CA.

significant increase in the gain of the eye move­ment velocities in the dark as compared to the gain obtained in the light conditions. In the dark trials, the gain was significantly higher when the subjects fixated on an imaginary target as opposed to a real target. There was no difference in phase among all conditions. This test could potentially serve for preliminary screening for the integrity of the vestibular system as it is noninvasive and of short duration. However, caution must be exer­cised in controlling various variables. Extensive nor­mative data are needed to properly assess this test as a screening tool for aircrews. Published 1996 by Elsevier Science Inc.

o Keywords - vestibulo-ocular reflex; high frequency; active rotation.

Introduction

The maintenance of visual orientation in a dy­namic environment is greatly enhanced by the fact tl1a1 the retinal is stabilized. One Clf the most seriolls threats to the stability of the retinal image comes from the observer's own head and body movements, and one of the most imp0l1ant stabilizing mechanisms is the vestibulo­ocular reflex (YOR), which generates compen­satory eye movements. While the visual system provides retinal stability at movements from 0 to about 2 Hz, the YOR becomes efficient at frequencies greater than 2 Hz. An absent or de­fective vestibulo-ocular reflex has potential for spatial disorientation in instrument flight. Recent

RECEIVED 11 October 1995; ACCEPTED 16 April 1996.

411

412

study (1,2) indicated that pathologic causes in­volving the perceptual consequences of vestibular anomalies have been implicated in a number of spatial orientation disturbances. Specifically, avia­tors with recurrent loss of spatial awareness were found to have a defective vestibulo-ocular reflex.

Unlike the visual system. the vestibular sys­tem l)f pilots is seldom investigated due to the lack of a practicaL reliable test and lack of normative data tor Currem::inicai assess-ments of vestibular integrity are time consuming and me primalily based on the exmnination of eye movements induced by caloric stimulation or in­duced by passive rotation at frequencies below 2 Hz. However, during natural movements, head movement frequencies between 1 and 4 Hz are most commonly experienced. In spite of consid­erable research, vestibular and equilibrium test­ing procedures are not standard or uniformly ac­cepted from one laboratory to another. An evaluation of diagnostic tests for vestibular func­tion was critically reviewed by the Working Group of the Committee on Hearing, Bioacous­tics and Biomechanics (3). The committee rec­ommends the Basic Vestibular Function Test Battery as the fundamental group of tests required for all clinical evaluations of vestibular integrity and for selection of groups in special occupations such as pilots. This test battery includes an EOG calibration, a saccade test, spontaneous and gaze­evoked nystagmus test, an ocular test, position­ing and positional test, and the caloric test. Such extensive testing would be extremely costly and time consuming for the initial medical screening of pilot candidates. What is needed is a prelimi­nary vestibular screening test that could indicate whether a full battery of clinical evaluation lS warranted. For example. a dynamic visual acuity test (oscillopsia test) is a useful screening test of visual-vestibular interaction (4). For equilibrium testing, the tandem gait test and sharpened Rom­berg tests in the dark are often used.

A number of attempts were made to investi­gate the VOR during head only (head-on-body) movements above 2 Hz. (5,6,7). The Vestibular Autorotation Test (V AT, Western Systems Re­search Inc.) has been used in numerous clinical assessments of the vestibular integrity of pa­tients suffering from a number of vestibular dis-

B. Cheung et al

eases such as Meniere's disease (8) and acoustic neuroma (9). It has also been used to monitor for cisplatin-induced vestibular toxicity (10), and for vestibulo-ocular abnormalities in pa­tients with panic disorder (1). The V AT test may be suitable as part of a series of screening tests for vestibular integrity to determine whether a full ENG battery is necessary. The apparatus, as described in the literature (8.9,10,11) pro­vides certain advantages over conventional clin­iCed It :s non-invasive 3.nd provides convenient assessment of the VOR in a rela­tively shol1 time. In addition, the high frequency stimulation also eliminates the problem of gain variability that is seen at frequencies below 1 Hz due to the voluntary manipulation by the pa­tient and the influence of visual stimulation. There remain, however, several unresolved issues. For example, head-only oscillation in time with an auditory signal also stimulates the neck recep­tors; testing can be done under various visual conditions which might alter the gain of the VOR especially at the lower frequencies. Also, some­times real targets are used and sometimes imag­inary targets. Exactly which procedures are best for this type of vestibular screening is not clear.

The current study attempts to assess the mer­its and efficacy of using the Vestibular Autoro­tation Test as a preliminary screening device for the integrity of vestibular function in volunteers from our laboratory. The group was composed of scientists, technicians, pilots, and graduate students. If proven viable, it could also be used in testing pilot-candidates, pilots who have ex­perienced recurrent or overwhelming spatial disorientation, or pilots who experience inability to focus on the instrument panel during turbulent IMC (instrument meteorological conditions).

Methods

Horizontal and vertical eye movements were reduced using electro-oculography during sepa­rate trials. That is, in any single trial, either the horizontal VOR or the vertical VOR was evalu­ated. Head velocities were recorded with cali­brated rotational velocity sensors. The head ve­locity sensor was independently calibrated at 100°1 s/volt. Eye velocity was calibrated from the first

Vestibular Influence on Head-shaking Test

2-s epoch of slow head and eye movements (frequencies below 1 Hz) by determination of the best least-squares linear regression fit of eye and head velocity after they were corrected for vergence and change of head position. Eye po­sition and eye velocity information were ampli­fied and digitized by a personal computer. The gain and phase of the input and output signals were computed by discrete Fourier analysis. Details of the computation of the gain and phase, algorithm employed, and signal-process­ing definitions have been documented prevI­ously by Fineberg and colleagues (12).

Subjects

A group of 10 subjects (8 men and 2 women) ranging from 25 to 59 years of age participated in the study. All of the subjects enjoyed excel­lent physical health; had no history of auditory, oculomotor, or vestibular disorder; and had no spontaneous nystagmus. Three of the subjects wore corrective lenses during the test. All of the subjects were able to visualize the target at the fixed distance of 145 cm with no difficulties.

The research protocol and procedures were reviewed by the DCIEM Human Ethics Com­mittee, and the subjects reviewed and signed the Human Experimentation Subject Volunteer Con­sent Form before their participation. The sub­jects also obtained medical approval to partici­pate from a DCIEM physician/medical officer. All subjects were instructed to strictly abstain from alcohol and cold medication for at least 24 h prior to testing, as alcohol and antihistamines (from cold medication) act on both the visual and vestibular systems.

Apparatus

The Vestibular Autorotation Test (V AT) is a computerized high-frequency test that uses a se­ries of audible tones sweeping through a graded frequency from 0.5 to 6.0 Hz as a stimulus for active head movements. It employs a 386 per­sonal computer for data collection, a V AT head unit with an eye movement preamplifier, a VAT head velocity sensor with a control unit, and an

413

eye/head interface cable. This apparatus has been used in hospital clinics without adverse ef­fects (9,11).

Procedure

All testing was conducted in the morning be­tween 0800 and 1200 hours. Each subject wore a padded crown (approximately 100 gm in weight) over his or her head with an attached head velocity sensor and the inputs for three surface skin electrodes. Three surface skin elec­trodes were placed on the subject's head with double adhesives: for horizontal VOR, one each beside the outer canthus of each eye and one reference electrode at the medial line of the forehead; for vertical VOR, the same reference electrode was used with one electrode each on top and below the same eye. The skin was pre­pared by thorough cleaning with an alcohol swab. Conductive electrode gel was placed be­tween the skin and the electrode cup. The sub­ject was seated in a firmly placed chair 145 cm in front of a wall-mounted target (a red light­emitting-diode powered by 3V DC cells) and was instructed to oscillate his or her head about the spinal axis (for horizontal VOR) smoothly in synchrony with an auditory tone generated by the V AT PC unit. This tone will sweep from 0.5 Hz to 6.0 Hz over a time course of 18 s. The in­structions to the subjects were, "Stare at the tar­get" and "Move your head smoothly from side to side in time to the tone." The procedure was repeated for a total of three trials with a resting period between trials. For vertical VOR, sub­jects were instructed to oscillate the head about the interaural axis. testing sessions were conducted on all SUbjects pnor to collect­ing data. These sessions included testing of the horizontal and vertical VOR in which subjects fixated on a wall-mounted target at a distance of ] 45 cm in the dark.

ExperiTnental Conditions

It is known that the sensitivity and gain of the VOR are highly plastic, influenced by many visual and nonvisual mechanisms and by envi-

414

ronmental factors such as lighting conditions, proximity of the visual scene or target, and the mental set or imagined percept that the subject chooses when measured during rotation in dark­ness. To determine the effective parameters for the VOR responses using the V AT apparatus, a previous study (12) compared the responses in the light with a real target and in the dark with an imaginary target only. In this study we inves­tigated in detail the influences of each of these two variables dn lhe \'estibulc­ocular reflex. The following five visual condi­tions were administered in random order to the subjects. Under each visual condition the sub­ject was tested 3 times on 3 different days to study the test-retest reliability within subjects. U sing condition l, the test and retest reliability check was expanded to 9, 18, and 22 times in 3 subjects, respectively.

Condition 1. Subjects fixate on the wall­mounted red LED target (145 cm in front of the subject) in a lighted room (Cl).

Condition 2. Subjects fixate on an imaginary target on a large blank white screen without any features (145 cm away and covering the subject's entire visual field) in a lighted room (C2).

Condition 3. After dark adaptation (30 min, in a darkened room with blindfold on), subjects fix­ated on a space-fixed imaginary target (C3).

Condition 4. After dark adaptation (30 min), the red LED target was illuminated for 3 seconds, subjects were instructed to fixate on the LED. The LED target was extinguished, subjects were instructed to continue fixating on where the tar­get was in the dark (with blindfold on in a dark­ened room) (Cc.\.).

Condition 5. After dark adaptation (30 min), the red LED target was switched on in the darkened room (the light in the laboratory was switched off), subjects were instructed to fixate on the il­luminated LED (C5).

It is known that changes occur in the corneo­retinal potential of the eye when the level of ambient illumination is altered (13), and these

B. Cheung et al

changes were shown to seriously affect electro­oculographic records. Therefore, the subjects were dark adapted for 30 min prior to calibra­tion and recording of eye movements, when the experiments were to be performed in the dark.

Data Analys is

Eye position and head velocity inforn1ation collected llsing the V AT were amplified and

a personal computer. veloc-ity was computed by the differentiation of eye position using a two-point central difference de­ri vative algorithm and calibrated from head sen­sor information from the first 2-s epochs of low frequency head and eye movements. The gain and phase of the input and output signals were computed from the last 12 s (2.0 Hz to 6.0 Hz) by discrete Fourier analysis (8). The gain and phase values were computed at 2,2.3,2.7,3.1, 3.5, 3.9, 4.3, 4.7, 5.1, 5.5, and 5.9 Hz, respec­tively. Gain is defined as the eye velocity ampli­tude divided by the head velocity amplitude. Phase is the time lag in degrees of the eye velocity in relation to the head velocity. An ideal VOR would be expressed as gain = 1 and phase =

180°, in which case equal-sized eye movements and head movements occur synchronous I y in op­posite directions. The statistical significance of the effect of the 5 conditions was tested by repeated measures, analysis of variance (Statview II) fol­lowed by post hoc test of Fisher's Protected Least Significant Difference. The multiple com­parison significance level was set at 990/0 (0.01).

Results

In this study, we attempted to investigate both the horizontal and vertical VOR using the V AT apparatus. However, we were not able to obtain a consistent and smooth vertical eye movement record during the preliminary trials when the subjects oscillated their heads actively about the interaural axis. Our discussion of the results will be limited to horizontal VOR. Fig­ure I shows a typical tracing of the relationship of eye position, eye velocity, and head velocity under condition C 1.

Vestibular Influence on Head-shaking Test 415

.30

20 10

0 0 w

0 -10

14 16 17 18

B

~r~V~NVWM~WW\fmNtrWN#JwN'

4 5 6 7 .'982 .1948 .2075 .1-990 .1987

8 9 10 SECONDS

11 12 13 14

Figure 1. Example of the relationship of eye position (A), eye velocity (B), and head velocity (C) of one of the subjects under condition 1 when the subject was instructed to fixate on an LED target 145 cm away in the lighted room and oscillate his head in time with the auditory signal.

Although the test procedure as described is relatively simple and noninvasive, from our preliminary study we found that training, prior to the collection of experimental data is often required for the subject to be able to oscillate his or her head in time with the given tone. In most subjects it takes up to 12 trials. In order to eliminate any training effects on the VOR, the training sessions and the experimental sessions were performed on separate days.

We were also interested in the reliability of data from trial to trial and the possible training effect within subjects. Three of tht

were exposed to 9, 18, and 22 trials, respec-6vely, under condition Cl (fixating on the space-fixed target in the light) on separate days. The mean gain of the V OR for each frequency in each of these subjects (and the corresponding standard etTOrs of the means) are shown in Fig­ure 2a, and the phase lag of the VOR from the head motion is shown in Figure 2b. With the in­crease in frequency, there is an increase in vari­ance of both the gain and the phase (the results are more variable at higher frequencies).

There was no correlation between the magni­tude of the VOR gain and the number of repeti­tions of the test in subject 1 under condition C 1 (illuminated target in illuminated room). There was in fact no correlation for any of the 11 fre­quencies tested. Subject 1 was tested 18 times on 18 different days. Similarly, there was no correlation between the magnitude of VOR gain and the number of repetitions of the tests in sub­ject 2, who was tested 22 times on 22 different days. There were no detectable training effects.

As stated , each subject was tested

plane to il1ustrate test-retest reliability within subject. The gain and phase measured for each subject for each frequency under each condi­tion was the average of the 3 trials. The mean gain and phase of all subjects versus frequency under the 5 different visual conditions are shown in Figure 3a (VOR gain versus fre­quency) and 3b (VOR phase lag versus fre­quency). When subjects were instructed to os­cillate their heads about the spinal axis in time with the auditory signal while observing a sta-

416

1.5

Mean 1 Gain

0.5

B. Cheung et al

A

••••••••• S1

--¢-- S2

-0- S3

4

Frequency (Hz)

Figure 2. Illustration of (A) the mean gain (maximum eye velocity/maximum head velocity at the various fre­quencies of head shaking and (8) the VOR phase values (time lag in degrees between eye velocity in relation to head velocity) at the various frequencies. The corresponding standard error of the means of the values at each frequency is shown by the vertical lines. S1 = subject 1,52 = subject 2,53 = subject 3. (continued)

tionary target in the light, the results were as expected; close to unity gain across all frequen­cies was observed.

Another limitation of the active head rotation testing using the V AT is that 4 subjects were unable to oscillate their heads beyond 4.7 Hz. This is partly due to the natural constraint of motion of the neck and, in some subjects, poor coordi­nation. Both of these limitations are subject-

specific. Data points missing at high frequencies are 7/150 (10 subjects,S conditions, 3 trials per condition) 36/150, and 76/150 for 5.1, 5.5, and 5.9 Hz, respectively. As a result the analysis was performed on 2 to 4.7 Hz where complete sets of data from each subject were obtained.

Statistical tests showed significant differ­ences between the VOR gains in the different visual conditions. Some subjects found it diffi-

Vestibular Influence on Head-shaking Test

250

225

(i) Q)

~ Cl Q)

e. Q)

200 If)

m .s:: 0..

c m Q)

:il

.L 175

T T 1 1

J.

B

1 I 1 1

I , "" .. "]

417

....•.... 81

J 1

--<>-- 82

-0-- 83

150

Frequency (Hz)

Figure 2. (continued)

cult to imagine a target. However, our results revealed that no significant difference between C3 (imaginary target in the dark) and C4 (sub­ject imagined where the target was after a target was shown for 3 s in the dark). This demon­strated that flashing a target prior to imagining the target made no significant difference in as­sisting the subject in the task. Statistical analy­sis reveals that from 2.0 Hz to 4.7 Hz, there is a significant difference between Cl (subject fix­ated on a real target in the light) and C3, and be-

tween C j and C4. The mean gains in C3 and C4 are both significantly greater than in Cl.

From 2.0 Hz to 4.7 Hz across subjects, the mean VOR gain obtained in C3 (fixation on an imaginary target in the dark) is significantly higher than the mean VOR gain in C2 (fixation on an imaginary target in the light) as shown in Figure 3a. Also, between 2.0 Hz and 4.7 Hz, the gain obtained from C3 (fixation on ilnaginary target in the dark) is greater than C5 (fixation on a real target in the dark) as shown in Figure 3a.

418 B. Cheung et al

A

1.5

Gain

0.5 -e-- c1

................. c2

·---0---- c3

----I::s,---- c4

- - -ffi- - - c5

1.5 2.5 3.5 4.5

Frequency (Hz)

Figure 3. Illustration of (A) the mean gain and (8) the mean phase values of all subjects across the frequency range of 2 to 6 Hz. (continued)

Table 1 summarizes the F-ratios and their re­spective P-values for significance between 2.0 to 4.7 Hz. The paired conditions at the higher frequencies (5.1 to 5.9 Hz) were not analyzed, due to the incomplete data set as mentioned ear­lier. Moreover, we observed that at higher fre­quencies the standard deviations of the gain were much larger. There were no significant dif­ferences between the phase relationships in any of the visual conditions (at any frequencies).

Discussion

The gain and phase characteristics of the VOR as induced by low frequency passive rota-

tion have been investigated extensively in both humans and nonhuman primates. In order to in­vestigate passive whole body oscillation at high frequencies (above 2 Hz), large, complicated, expensive equipment must be employed and nu­merous technical difficulties have to be over­come. As a result, active or passive head-only rotational tests have been advocated as methods of assessing the VOR at high frequencies.

Few attempts have been made to measure the VOR gain and phase characteristics during ac­tive head rotation at frequencies of 2 to 6 Hz. With the subject's head firmly fixed in a metal frame fitted with a dental bite board (5), sub­jects were instructed to shake their heads while observing a stationary target or in darkness, a unity gain of 0.9 to 1.1 was observed at all but

Vestibular Influence on Head-shaking Test

250

225

Phase (degrees)

200

175

419

8

-.-- c1

................. c2

----0---- c3

----il---- c4

- - -ffi- - - c5

150 -t----.---.-----,---.---.,..----,-----, 1.5 2.5 3.5 4 4.5

Frequency (Hz)

Figure 3. (continued)

the highest frequencies. Between 5 and 6 Hz, a small decrease in the gain was observed.

Our results are comparable to earlier reports on active head-only rotation at high frequency. ] ell and coll eagues (6) reported that with the head unrestrained and wheE an eartb-fixed ta"­get was visible, the mean VOR gain in the light ranged from 1.13 to 1 Al between 2 to 5 Hz. By comparison, using the V AT apparatus under similar visual conditions, we obtained a mean gain of 1.03 to 1.04 between 2.0 to 5.9 Hz. Across all subjects, the mean gain obtained when the visual target is visible (condition C 1) is significantly lower than the mean gain ob­tained in darkness (conditions C3 and C4). Ear­lier work by Keller (14) reported similar find­ings in monkeys. Under sinusoidal oscillation, the VOR gain in the dark increased steadily un-

til it exceeded unity at about 2 Hz, and maxi­mum VOR gain was reached at about 4 Hz in the dark with a mean gain of l.3. However, our finding is in direct contrast to previous findings hy Fineberg and colleagues (12). The authors

thai. the same there

was no significant difference in the gain ob­tained from head movement while fixating a stationary wall mounted target and the gain ob­tained from head movement in darkness with an imagined stationary target. Also, Jell and col­leagues (15) reported that at frequencies above 1 Hz, horizontal VOR gain in response to active oscillation approached a value between 0.7 and 0.9, regardless of whether the visual target was head-fixed or earth-fixed, real or imagined.

One of the possible explanations for the above result could be the inability of the sub-

420

Table 1. Summary of the F-ratio and their respective P-values of significance difference between

conditions C2 and C3 and conditions C3 and C5

Frequency F-ratio p-

value

2.0 Hz 7.833 0.0001 2.3 Hz 8.191 0.0001 2.7 Hz 8.3657 0.0001 3.1 Hz 8.904 0.0001 3.5 Hz 7.522 0.0002 3.9 Hz 5.508 0.0014 1..6 l-jz 5.605 0.0013 4.7 Hz 5.166 0.0022

jects to maintain the same state of vergence as that maintained when a real target is available. If vergence is not maintained, and if the dis­tance of the imaginary target decreased (from 145 cm) in the dark, extra angular distance would have to be traversed to maintain ver­gence on the target (16,17). As a result, the ini­tial calibration during the first 6 s of 0.5-Hz head movement, will be incorrect, and the VOR gain would increase. In a preliminary study we have examined this issue further on 6 available subjects who participated in the original study. U sing a binocular eye tracking system, sam­pling at 120 Hz and with a resolution of 0.1° (EL-MAR Inc. Series 2000), we asked the sub­jects to fixate on the same LED light placed at 145 cm straight ahead for 20 s. After the light was extinguished, the subject was asked to maintain fixation at the point where the LED was for another 20 s. The same amounts of training were given as previously administered in the experiment. The subjects were encour­aged to fixate at the imaginary target during the trial. Using the measured interpupillary distance of the subjects, the recorded position of each eye, by simple geometry, we compared the ver­gence distance achieved when the LED light was on and when the LED light was off. Our re­sults showed that for the 20 s when the light was on, there were negligible movements of both eyes and the movements were conjugate. The average vergence distance was 145 :±: 0.5 cm. When the light was off, there were relatively more oscillations. During the first 16 s, the eyes move conjugately in the same direction and by the same amount. That is, from the initial posi-

B. Cheung et al

tion, if the left eye moves to the right by 0.2°, the right eye also moves to the right by 0.2°. As a result, the position of the imagined target would have shifted laterally on a plane at a fixed distance from the subject. The vergence distance was calculated to be 145 :±: 1.3 cm. The observed conjugate movements of the two eyes deteriorate after 16 s. It appears that with encour­agement and motivation, subjects could hold their gaze and focus at a fixed distance at least /~)I' a snort Lime related to the duration of retinal after-image). The duration of the VAT procedure lasts 18 s and covers the frequency of 2.0 to 6.0 Hz. Our results that were reported to be significant were from 2.0 to 4.7 Hz, well within a 16-s period. However, the possibility still exists, as we were not able to technically ver­ify the vergence distance during the original study. It warrants a full scale study.

It is assumed that high frequency testing has the advantage that visual ocular control (smooth pursuit) is not active above 2 Hz, thus removing visual influences of real target or imagined tar­get. Contrary to this assumption, we found here that the interactions of VOR with pursuit reduce the gain of the VOR during visual fixation. This finding is in disagreement with that of previous studies. Fineberg and colleagues (12) used the same apparatus and reported that there were no significant differences in the 2- to 6-Hz range between subjects tested in the light compared with those tested in the dark. On the other hand, J ell and colleagues (6) used a different tech­nique and reported that from 0.2 to 3.9 Hz, the VOR gain was slightly but significantly greater with visual fixation of space-fixed target in the light, compared to fixation of imaginary target in the dark. Recent data from Furman and Dur­rant (18) indicated that the \fOR gain with vi­sual fixation at 1 and 2 Hz only was slightly higher than with no fixation.

Two factors were involved in conditions C1 and C4, namely, the availability of a space­fixed visual target and the availability of light during testing. Therefore, further attempts were made to investigate each of the factors sepa­rately. The above findings-that the VOR gain was reduced when the visual target was visible­were further confirmed by the paired difference between conditions C3 and C5 from 2 to 4.7 Hz.

Vestibular Influence on Head-shaking Test

Our results indicate that the mean VOR gain in condition C3 (dark with imaginary target) was greater than C5 (dark with real target).

The other paired difference that was signifi­cant was between conditions C3 and C2 when the visual target was not available. Although the subject was instructed to fixate on an imaginary target on a blank screen with no visible features, in C2, the presence of light reduces the gain of the VOR measured. The mean gain of the VOR between 2.0 to 4.7 Hz. is consistently higher when the test was performed in the dark than in the light when the subjects were instructed to fixate on an imaginary space-fixed target under both conditions. High frequency gains differ with visual tasks, as do the gains at lower fre­quencies. Since the VOR gains in the dark are greater than one, it seems natural that the effect of a fixed visual target would be to "improve" the gain by decreasing it towards one.

The results indicated that the ability to mod­ulate the gain of the vestibulo-ocular reflex does not depend entirely on the smooth pursuit, sys­tem. This finding is in agreement with that of Derner and colleagues (19) that there is signifi­cant visual-vestibular interaction during active head movements at frequencies up to 6 Hz. It suggests a role for other mechanisms besides in­teraction of VOR and pursuit system during head-only rotation while fixating on a space­fixed target. Other centers must modulate eye velocity so that it is appropriate to the subject's choice of a frame of reference, whether a target is visible or not.

The possibility of a contribution by the mo­tor program for head movements to the fine reg­ulation of compensatory eye movements has previously been ignored. A recent study by Hoshowsky and colleagues (20) indicated that there is a significant ditIerence between the

421

mean gain obtained during active high fre­quency (2 to 4 Hz) head-only oscillation as compared to whole body oscillation. Similarly, the mental set of the subject when viewing a vi­sual target, real or imagined, can have profound effects upon the response.

Our prelill1inary attempts at recording verti­cal eye movement using the V AT were not suc­cessful. The records included a great deal of noise and artifacts, and the results obtained were not consistent from trial to trial within subjects. We attributed these artifacts to the fact that the EOG technique is very sensitive to lid closure (due to eye blink) and other sources of physio­logical and environmental noise and interference. EOG artifacts during vertical eye movements have been known to have the same time course as the upper eyelid movement (21). Our inabil­ity to obtain consistent and analyzable record is probably directly attributable to this factor. In addition, there is an inherently poor signal-to­noise ratio, given the small amplitude of eye movements generated by head-only rotation.

In conclusion, our results provide additional information about influences on the VOR dur­ing active head-only rotation. This study con­firmed that the V AT apparatus is well tolerated by subjects, however, based on our experience, it is time consuming because of the training that is required for the subject to learn to oscillate the head in time to the relatively high-frequency tone. This requirement for extensive practice sessions will probably reduce the usefulness of this technique in aircrew screening. Also, it is important to control carefully the visual condi­tion and the mental activity during the test. Ex­tensive normative data using this test in asymp­tomatic would be needed. and results from subjects Known (0 be tive would also be needed.

defcc-

REFERENCES

1. Clark .TB, Angus RH. Spatial disorientation and dys­function of orientation equilibrium reflexes: aeromedi­cal evaluation and considerations. Aviat Space Environ Med. 1992;63:914-8.

2. Rupert A, Guedry FE, Clark J. Medical evaluation of spatial disorentiation mishaps. AGARD Conference Proceedings 532 April 27-May 1, 1992. Aircraft acci­dents: Trends in Aerospace Medical Investigation Techniques, Cesme, Turkey: presented at the Aero-

space Medical Panel Symposium, 27 April-l May, published by AGARD printed by SPS, Loughton Essex, 1992;AGARD, Loughton, Essex, 57-J-57-5.

3. Von Gierke HE (Chair). Evaluation of tests for vestibu­lar function. Committee on Hearing, Bioacoustics, and Biomechanics. A viat Space and Environ Med. 1992;63 (2, suppl.):AI-A34.

4. Barber HO. Vestibular neurophysiology. Otolaryngol Head Neck Surg. 1984;92:55-8.

422

5. Tomlinson RD, Saunders GE, Schwarz DWF. Analysis of human vestibulo-ocular ret1ex during active head movements. Acta Orolaryngol. 1980;90: 184-90.

6. Jell RM, Guedry FE JR, Hixson C. The vestibulo-ocu­lar ret1ex in man during voluntary head oscillation under three visual conditions. A viat Space Environ Med. 1982;53(6):541-8.

7. Gauthier GM, Piron JP, Roll JP, Marchetti E, Martin B. High-frequency vestibulo-ocular reflex activation through forced head rotation in man. Aviat Space Environ Med. 1984;55(1): 1-7.

8. O'Leary DP, Davis LL. Vestibular autorotation testing of Meniere's disease. Otolaryngol Head Neck Surg. 1990; I 03:66-71.

GP. DP '/cstibular :mtoro-tation test asymmetry analysis uf acollstic neuromas. Otolaryngol Head Neck Surg. 1991: l04( l): 103-9.

10. Kitsigianis GA, O'Leary DP, Davis LL. Vestibular autorotation testing of cisplatin chemotherapy patients. Adv Otorhinolaryngol. 1988:42:250-3.

11. Hoffman DL, O'Leary DP, Munjack DJ. Autorotation test abnormalities of the horizontal and vertical vestib­ulo-ocu1ar reflexes in panic disorder. Otolaryngol Head Neck Surg. 1991; 110(3):259-69.

12. Fineberg R, O'Leary DP, Davis LL. Use of active head movements for computerized vestibular testing. Arch Otolaryngol Head Neck Surg. 1987;113:1063-5.

13. Gonshor A, Malcolm R. Effects of changes in illumina-

B. Cheung et al

tion level on electro-oculography (EOG). Aerospace Med. 1974;2(2): 138-40.

14. Keller EL. Gain of the vestibulo-ocular ret1ex in monkey at high rotational frequencies. Vision Res. 1977;18:311-5.

15. Jell RM, Stockwell CW, Turnipseed GT, Guedry FE. The int1uence of active versus passive head oscillation, and mental set on the human vestibu1o-ocular reflex. Aviat Space Environ Med. 1988;59: 1061-5.

16. Virre E, Tweed 0, Milner K. Vilis T. A reexamination of the gain of the vestibuloocular reflex. J Neurophys­iol. 1986;56:439-50.

17. Huebner WP, Polaski WH, Reschke MF, Bloomberg JJ. Geometric adjustments to account for eye eccentricity in processing horizontal and vertical eye and head movement datil. J Vestib Res. 1995::5(4):299-322.

! is. Furman .1M, Durrant JD. Head-only rotational testing: influence of volition and vision. J Vestib Res. 1995;5(4):323-9.

19. Demel' JL, Oas JG, Baloh RW. Visual-vestibular inter­action during high-frequency, active head movements in pitch and yaw. Ann N Y Acad Sci. 1992;656:832-5.

20. Hoshowsky B, Tomlinson D, Nedzelski J. The horizon­tal vestibulo-ocular reflex gain during active and pas­sive high frequency head movements. Laryngoscope. 1994; 104: 140-5.

21. Barry W, Melville Jones G. Int1uence of eye lid move­ment upon electro-oculographic recording of vertical eye movements. Aerospace Med. 1965;36:855-8.