266
Electroencephalography and clinical Neurophysiologv, 1985, 6 2 : 2 6 6 - 2 7 6 Elsevier Scientific Publishers Ireland, Ltd.
OBSERVATIONS U P O N THE EVOKED R E S P O N S E S TO NATURAL VESTIBULAR STIMULATION J D . H O O D 1 and A. K A Y A N 2 Medical Research Council Neuro-Otology Unit, Institute of Neurology,, National Hospital. Queen Square, London W C I N 3BG ( U.K.)
(Accepted for publication: February 28, 1985)
Summa~ Repetitive rotational stimuli simulating natural head movements have been applied to the study of the vestibular evoked response in normal subjects and 12 patients with complete loss of vestibular function. Special precautions were taken to eliminate all possible sources of artefacts, in particular, all eye movements were restrained by requiring the subject to fixate upon a target light attached to the rotating chair throughout the course of the test. With a stimulus of 2 sec duration the typical response took the form of a slow negative wave with a mean peak amplitude of approximately 24/~V and maximally recorded from the vertex. It was characteristically absent in the patient group. Occasionally, both in normal subjects and patients it was preceded by a long latency complex thought to be non-vestibular in origin. Tests carried out both in total darkness and in the light show a statistically significant increase in the potential in the latter condition indicating an influence of the optokinetic effect exerted by the visual surround. Further studies have explored the phase changes brought about by varying the amplitude and duration of the stimulus. These have revealed certain parallels in the results of recent animal experimental studies. Keywords: t,estibular et,oked response - - rotation
absent ~,estibular function --. optic' fixation - - corneo-retinal potential
Although auditory and visual evoked potentials have now found extensive clinical application, progress in the recording of vestibular evoked potentials has been disappointingly slow and to date, we have been unable to trace more than 15 papers on the topic in the world literature. The lack of enthusiasm is not difficult to understand. Whereas both the auditory and the visual systems are ideally suited to the application of the short duration repetitive stimuli demanded by modern averaging techniques, the mechanics of the vestibular system have long time constants which preclude this kind of approach. In addition, no single clearly defined cortical area associated with the vestibular system has been identified. A particular problem, however, that needs to be
I Address for correspondence: Dr. J.D. Hood, M R C NeuroOtology Unit, National Hospital, Queen Square, London W C I N 3BG, U.K. 2 Supported by the Wellcome Trust. Present address: Royal National Throat, Nose and Ear Hospital, Gray's Inn Road, London WC1, U.K.
t,erte r
given special consideration is the difficult one of ensuring that artefacts do not arise either from electrode movement caused by rotation of the subject or by spurious potentials induced by the electro-magnetic field of the driving motor. Of even more serious concern, however, are the eye movements induced by any rotational stimulus by way of the vestibulo-ocular reflex (VOR) arc, a matter of some importance since Hillyard and Galambos (1970) have provided clear enough evidence of contamination of the contingent negative variation (CNV) by eye movement artefact under conditions in which eye movements were left unrestrained. In the report of the committee on Publication Criteria for Studies of Evoked Potentials in Man (Donchin et al. 1977) the following pertinent comment appears: 'There is a potential difference of about 100 mV between the aqueous humor (positive) and the retina. Any movement of the eyeball causes a change in orientation of the potential field which affects the scalp electrodes (certainly as far as the vertex) in proportion to their distance from the eyes. The possible contamination from eye movements should be a major
0168-5597/85/$03.30 © 1985 Elsevier Scientific Publishers Ireland, Ltd.
VESTIBULAR EVOKED RESPONSE
267
c o n c e r n to all EP investigators a n d the measures t a k e n to deal with this p r o b l e m should be considered in a n y p u b l i s h e d report.' Despite this, the m a t t e r has been given scant a t t e n t i o n in a n y of the studies of vestibular evoked potentials which have as yet appeared a n d n o serious a t t e m p t has been m a d e to eliminate this source of artefact. I n the investigations to be described particular p r e c a u t i o n s have b e e n taken to m i n i m i s e all possible sources of artefacts of mechanical, electrical a n d physiological origin. As a result we feel reas o n a b l y c o n f i d e n t that the responses described in this paper can be a t t r i b u t e d to nervous activity e m a n a t i n g from the vestibular system following the application of rotational stimuli.
Subjects Subjects were d r a w n from a pool of 15 healthy adult volunteers; 10 males a n d 5 females whose TABLE I Patients with absent (or greatly reduced) vestibular function. Case
Age
Sex
Aetiology
Duration of symptoms
1. AH 2. AB 3. RS 4. JS
62 34 48 53
M M M F
2 months 22 years 2 years 20 years
5. JF
55
M
6. PR
33
M
7. AQ 8. CN 9. PH 10. BC
50 78 34 27
F M F M
11. DW
21
M
Ototoxicity Ototoxicity Ototoxicity Meningitis and ototoxicity Meningitis and ototoxicity Meningitis and ototoxicity Ototoxicity Ototoxicity Post-infective Unknown (? migraine) Bilateral acoustic neurofibroma (R. removed Von Recklinghausen's disease) Bilateral acoustic neuroma (L. removed Von Recklinghausen's disease)
12. RT
25
M
3 years 3 years 5 months 6 months 9 years 10 years 19 years
ages ranged between 21 and 59 years with an average age of 36.5 years ( + 1 1 . 8 ) . I n a d d i t i o n tests were carried out u p o n 12 patients found, on neuro-otological e x a m i n a t i o n , to have bilateral a b o l i t i o n or gross reduction of vestibular function. The suspected aetiology is given in Table I. N i n e patients had n o response to caloric or rotational stimuli a n d in the r e m a i n i n g three only m i n i m a l responses could be detected. All patients comp l a i n e d of i m b a l a n c e which was exacerbated by oscillopsia in six. Deafness was present in all but one patient.
Method The rotating chair used in these studies incorp o r a t e d a direct drive, servo-controlled, Contraves a n d Goertz t u r n t a b l e system with a peak torque of 80 ft lb a n d peak acceleration of 1000°sec -2. It was completely silent in operation. The stimulus profile was p r o g r a m m e d by a K r o h n - H i t e function generator a n d in most of the studies to be described took the form of a raised cosine with a d u r a t i o n of 2 sec repeated at intervals of 12 sec with the stimulus parameters listed in the first line of T a b l e II. It simulated the form of a n o r m a l head m o v e m e n t a n d because of its smooth transition was designed to m i n i m i s e as far as possible a n y m o v e m e n t o f electrodes. Additionally, in further studies similar stimuli with d u r a t i o n s of 1 a n d 0.5 sec a n d parameters, as listed in T a b l e II were used. U s u a l l y 30 or occasionally a m i n i m u m of 15 stimuli were delivered with clockwise a n d anticlockwise chair rotation. As shown in Fig. 1, the subject rested his chin TABLE I1 Stimulus parameters for 2 sec, 1 sec and 0.5 sec raised cosine rotation. Duration Peakacceleration/ Peak velocity Chair of rotation deceleration displacement 2 sec
7 years 1 sec 0.5 sec
151 ° sec -2
192° sec -2 173° sec 2 205° sec -2 152° s e c - 2 200° s e c - 2
88° sec -1
120°
51° sec 1
30°
23° sec -1
7.5°
268
J.D. HOOD, A. KAYAN Frame for E l e c t r o d e Connections HE
;t
Target Pre-amplifiq
Light
S e a t Restr=
Fig. 1. Schematic diagram of rotating chair.
in a chin rest with his head inclined forward 30 ° and in this position a padded V-shaped head support was brought forward and secured on the occiput thus comfortably immobilising the head. Immediately to his front he was presented with a structured fixation target subtending a visual angle of 0.5 ° consisting of an L E D housed in the base of the chair and conveyed by way of a fibre optic system. Fixation upon this target was maintained throughout all test procedures and monitoring of eye movements by means of DC electro-oculography confirmed that any eye movements that may have been induced by rotation of the chair were below the limits of resolution of the recording system. In practise this means that they could not have exceeded 0.5 ° . When occasional inadvertent blinks intervened the relevant epoch was rejected. An accelerometer attached to the head confirmed that head and chair movements were synchronous. The chair was housed in a light-tight room so that testing could be carried out either in total darkness or with the room illuminated. In the latter event the vestibular stimulus was accompanied by an optokinetic stimulus evoked by the
images of the room background traversing the retina of the virtually stationary eyes with fixation maintained upon the target light. Ag/AgC1 stick-on cup electrodes of 10 m m diameter were used for the recording of evoked responses. The inter-electrode impedance was measured prior to and at intervals throughout the test procedures. It was usually of the order of 2 k~2 and never exceeded 4 kQ. The usual electrode configuration adopted was vertex-active, tip of nose reference and nasion-ground though in certain studies the vertex was referred to one or other mastoid. Additionally the scalp topography was explored using electrode placements according to the international 10-20 system. Following convention, upward deflection indicated negativity and downward positivity. The amplifier had an input impedance of 101~ ~2 with a low frequency cut-off of 0.06 Hz and a high frequency cut-off of 90 Hz each with a slope of 6 dB/octave. Responses were observed on-line and simultaneously recorded for further analysis on a Hewlett Packard FM data recorder with a tape speed of 15/16 c m / s e c and bandwidth DC to 312 Hz. Signal averaging was carried out upon a Solartron 1200 signal processor at a sampling rate of 85 kHz and amplitude resolution of 2.1 t~V. For calibration purposes a floating signal generator with a peak-to-peak amplitude of 100 t~V was connected to the input of the amplifier with a fixed gain and the transfer function computed on the Solartron signal processor. Peak amplitudes were measured with reference to a baseline defined as the average over the 750 msec epoch preceding the onset of the stimulus.
Results In Fig. 2 is shown a typical averaged response from a normal subject to thirty 2 sec stimuli with the ground electrode on the nasion and the vertex electrode referred to the tip of the nose. The gross morphology of the response comprises a domeshaped slow negative variation with a mean peak amplitude of 23.52 /,V (+5.33). It was nearly always clearly defined following the first 10 stimuli and its duration matched that of the stimulus. Both clockwise and anti-clockwise rotation evoked
VESTIBULAR EVOKED RESPONSE
269
COUNTERCLOCKWISE ROTATION
CLOCKWISE ROTATION /
---\
f - /g*~.
c z ~ P ' ~ C ~ w" ~ ,
~,"
~i~ "~
1 see
Fig. 2. Typical slow negative potential in response to 2 sec stimulus (St) recorded from vertex/nose in the presence of optic fixation in the light and in the dark. Note the preceding initial long latency potentials (30 responses averaged). Arrow indicates stimulus onset. The chair signal in this and subsequent figures refers to velocity.
the same response. Occasionally it was followed by a slow positivity which was ill-defined and conformed to no particular pattern. However, in 45% of the recordings in darkness and 65% in the light, a biphasic wave preceded the major component of the response. This initial wave complex consisted of a negative wave followed by a positive wave with average latencies of 249 msec _+ 16.2 and 404.6 msec +_ 37.2 respectively. It showed considerable inter- and intra-subject variability being present in one experiment and absent in another during the same session for any given subject.
Electrode placement Scalp topography of vestibular evoked potentials. Averaged slow potentials evoked by the 2 sec rotational stimulus were identified at a number of electrode locations. As mentioned earlier the mean amplitude at the vertex referred to the tip of the nose in 10 subjects was 23.52 /~V (_+5.33). In a further group of 5 normal subjects recordings from the vertex referred to the mastoid gave a mean amplitude of 23.4 ~V (_+ 3.9), similar values being obtained with the chin as reference. Mastoidmastoid, mastoid-chin and mastoid-nose recordings elicited no obvious response indicating that the mastoid was not an active site. The mean amplitude values obtained from other locations using the 10-20 electrode placement and reference electrode at the tip of the nose or the mastoid were as follows: mid-frontal region (Fz) 8.9/~V, central regions (C3 and C4) 6.7 ~V, parietal
regions (P3 and P4) 7.6 /~V, posterior temporal regions (T5 and T6) 4.9 ~V and mid-occipital region (Oz) 4.5/IV. As can be seen from Fig. 3 the responses at the vertex were markedly greater than those at any other electrode location. Evoked potentials recorded from the lateral frontal regions (F3 and F4) were either absent or poorly formed. There was no significant amplitude difference in the left right distribution or in respect of ipsilateral or contralateral rotation. By way of investigation of any influence the corneo-retinal potential might have had upon the responses, recordings were carried out of horizontal electro-oculography (EOG) recorded from nose and left outer canthus and nose and right outer canthus together with concurrent recordings from the vertex and F3 and F4. Similar responses were obtained between F3 and F4 and right eye and left eye and in consequence only right eye and F4 are displayed in Fig. 4 which represents the average of 30 responses. No obvious contamination by eye movements can be detected. Although there were no differences in the wave form and amplitude of the potentials recorded
CLOCKWISR EOTATION
(DARK)
St.
1
Cz-Fz
C P
z 4
~ ~ ,
O i 1le¢
~ M
V
Fig. 3. Scalp distribution of slow negative response to 2 sec stimulus (St) in the dark (normal subject). In monopolar recordings reference electrode was on tip of nose (30 responses averaged). Arrows indicate stimulus onset.
270
J.D. HOOD, A. KAYAN CLOCKWISE ROTATION
COUNTER CLOCKWISE ROTATION
/-\\
~J
Rt Eye
ilOvV Fig. 4. Responses from vertex, F4 and right outer canthus referred to nose. Similar responses were obtained from F3 and left outer canthus. from the vertex referred either to the mastoid or the tip of the nose, the latter was used routinely in the experiments which follow since it was considered less susceptible to contamination by corneoretinal potentials evoked by inadvertent horizontal eye movements.
Habituation Greiner et al. (1967), using a similar stimulus profile with a duration of 4 sec and maximum amplitude of 170 °, reported a progressive reduction in the evoked response with almost total extinction following the application of some 30-35 stimuli. This they attributed to habituation, a phenomenon which is now well recognised as being instrumental in bringing about a marked reduction in the nystagmic response induced by repeated rotational stimuli. This, of course, is an important consideration in respect of averaging techniques applied to the evoked response since in this event it is to be expected that in contrast to auditory or visual evoked responses, increasing stimulus numbers instead of enhancing the response profile would tend to degrade it. As it happens we can find no evidence in our own studies of such an effect. Seven subjects were rotated as outlined earlier using the 2 sec stimulus profile with the chair rotating clockwise and anticlockwise. Thirty stimuli in all were delivered in each direction with 12 sec rest interval between each stimulus and recording from vertex referred to the tip of the nose. Averaging was carried out
upon sequential groups of 10 stimuli together with the total number. Typical results from one subject are shown in Fig. 5. No significant differences were found in the responses to clockwise and anti-clockwise rotation either in the light or the dark. In addition, repeated measures analysis of variance (RMAV) revealed that there was no significant difference between the averaged responses to the 1st, 2nd and 3rd successive 10 stimuli or the total number. In consequence we feel reasonably confident that habituation is not a complicating factor which needs to be taken into consideration in respect of the particular stimulus profile applied in these studies.
Effect of stimulus parameters and visuo-vestibular interaction upon evoked responses The following stage of our study served a twofold purpose. First as an enquiry into the effects of changing the stimulus parameters upon the amplitude and phase of the response and second, the effect of adding a visual to the vestibular stimulus. Three stimuli were used with durations of 2, 1 and 0.5 sec and parameters as listed in Table II. Ten normal subjects took part and were tested over 3 sessions each lasting about 1.5 h. At each session a different stimulus was used. First the subject was rotated in total darkness with fixation upon the target light maintained throughout the test. Thirty stimuli at intervals of 12 sec were delivered in a clockwise and then in an anti-clockwise direction. CLOCKWISE ROTATION
St
COUNTERCLOCKWISE ROTATION
~
z~
ww°a:Cz~'~'~,,-,.,,,~
~
,
o <
[20pv I sec
I se¢
Fig. 5. Averaged vertex response to sequential groups of ten 2 sec stimuli together with total average. Rotation in dark in presence of optic fixation.
V E S T I B U L A R E V O K E D RESPONSE
271
This procedure was then repeated with the laboratory well illuminated so that during the course of sustained fixation upon the target, rotation of the subject induced a visual stimulus produced by the traversing of the images of the environment across the effectively stationary retina. Monitoring of eye movements by means of electro-oculography revealed that during the course of the 2 sec stimulus, no eye movements were detectable either in the horizontal or vertical plane. With the 1 sec and more particularly the 0.5 sec stimulus, occasional movements of small amplitude were observed. Typical averaged responses in one subject for the 3 stimuli are shown in Fig. 6 where it will be seen that the wave form changes markedly as the stimulus duration is reduced. Amplitudes measured in microvolts from the baseline to the peak of the slow negative wave are displayed graphically in Fig. 7 where open circles refer to responses in the light, closed circles, responses in the dark. The first point of note is that both in the light and the dark the amplitude decreases with decrease in the duration of the ;LOCKWlSE ROTATION IN THE DARK WITH FIXATION
St1 __
i20p v
Cz
St2 - - - -
i20pV
0'.5
1j see
Fig. 6. Vertex response to 2 sec (Stl), 1 sec (St2) and 0.5 sec (St3) stimuli. Note phase advance of response with respect to velocity of 2 sec stimulus (15 responses averaged). Arrows indicate stimulus onset.
AMP, microvol 40_
o Liqht • Dark
o
o ___O___
30.
o 8 o°° - , i l -
8
20_
8o 1 ~ 10_
--o.-
•
o
--6--
2 Sec
1 Sec
Duration
0.5 Sec
of r o t a t i o n
Fig, 7. Amplitudes of vertex response to 2, 1 and 0.5 sec stimuli in 10 normal subjects with optic fixation in the light and in dark. Clockwise and counter-clockwise values have not been separately symbolled since they were found not to be significantly different. Sofid lines, mean amplitude; broken lines, 1 S.D.
stimulus. Repeated measures analysis of variance shows that the difference between the 2 sec and 1 sec stimulus is significant at the level of P = 0.0009. N o significant differences are apparent in the responses to clockwise and anti-clockwise rotation in the light or in the dark. The second point of some importance is the difference in amplitude with all three stimuli between the responses in the light and in the dark. These differences are significant at the level P = 0.0012. Although it is not immediately apparent on direct examination of the respective responses this demonstrates quite clearly that the addition of a visual stimulus makes a small but significant contribution to the overall response and more importantly, that visuo-vestibular interaction is detectable in the evoked response. Phase relationships. It is evident from a cursory examination of the tracing shown in Fig. 6 that the phase relationships between response and stimulus change with the duration of the stimulus. Interestingly, similar changes are evident in extracellular responses at vestibular nuclei (Waespe and Henn 1977) and thalamic levels (Bi)ttner and Henn 1976) in monkeys in comparable test situations. In our
J.D. HOOD,
272
studies we have applied 2 forms of analysis to the examination of phase relationship. (1) Cross-correlation analysis of the velocity signal of the stimulus and the vertex response in the time domain using a Solartron 1200 signal processor. Negative values afforded phase lead and positive values phase lag in milliseconds. (2) Transfer function analysis of the phase angles at the dominant frequencies of the raised cosine stimulus. This provided added confirmation of the results established at the first analysis. Unfortunately, identifiable responses with the 0.5 sec stimulus were either absent or too small to provide any meaningful data. The phase relationships of the 2 sec and 1 sec stimulus, however, are displayed in Fig. 8 and show that both in the light and the dark the response is phase advanced on average by some 160 msec whereas in the case of the 1 sec stimulus the phase is slightly delayed. The results of the cross-correlation analysis in terms of time differences between signal and stimulus velocity showed that the differences are highly significant at the level P = 0.0004. There were no significant differences between rotation in the light and the dark but surprisingly perhaps, significant time differences in phase angles were apparent between clockwise and anti-clockwise rotation ( P = 0.0481).
TABLE
A. K A Y A N
II1
Stimulus parameters for 2 sec raised cosine rotation * at 90 °, 60 ° a n d 30 ° s e c -
1
Duration of chair rotation
Peak velocity
Peak acceleration/ deceleration
Chair displacement
2 sec
90 ° sec - 1
112 ° sec 120 ° sec
120_130 °
2sec
60 ° s e c
]
72 ° s e c 2 76 ° s e c - 2
90 °
2sec
30 ° s e c
1
35 ° s e c 2 35 ° sec - 2
45 °
2 2
* Rotation in the dark with fixation.
Effect of change in peak velocity In the foregoing the durations of the stimuli were changed with peak angular acceleration remaining reasonably constant. It is of interest,
40
0 > 0
30
0
E o LIGHT • DARK
-400
"10
20
.
-300
T
E "",,, ,
;
10
-100
.--•
0 .100 *200
I
i
2
1
i
0.5 sec
0 I
i
2
0.5
sec
DURATION OF ROTATION
3 0 ° S "1
6 0 ° S "1
9 0 ° S'l
PEAK VELOCITY
Fig. 8. Cross-correlation analysis of time differences in milliseconds of slow vertex potentials to 2, 1 a n d 0.5 sec stimuli from 10 subjects as in Fig. 7. Symbols indicate mean values, bars 1
Fig. 9. Amplitude of vertex response to 2 sec stimuli with peak velocities of 30, 60 a n d 90 ° sec -~. Included are values for each
S.D. Negative values, phase advance; positive values, phase lag re-stimulus velocity.
lines, mean amplitudes; broken lines, 1 S.D.
subject for clockwise and counter-clockwise rotation. Solid
VESTIBULAR EVOKED RESPONSE R.T.
CLOCKWISE
273
ROTATION
(DARK)
LIGHT
R
~V
.
I
_
COUNTERCLOCKWISE
R
L
°
o
°
R: C l o c k w i s e rotation L: C o u n t e r c l o c k w i s e rotation
ooo
30-
20-
L
o
40-
st.
DARK
Am
o
o6o
6'
-%o -
i
•
eo
%
ROTATION
St.
10 ¸
'i o D
I
o
DIIOIQ
[3191{3
IN THE LIGHT 1 sec J.F.
CLOCKWISE
IN DARK ROTATION
(DARK)
St.__
COUNTERCLOCKWISE
ROTATION
St
cz
1
I
lOpV
1 sec * Fig. 10. Absence of slow vertex response to 2 sec stimulus in 2 patients with absent vestibular function (cases 5 and 12 - Table I) (30 responses averaged).
therefore, to consider the effect of changes in peak velocity while maintaining the duration constant at 2 sec. Three stimulus wave forms were used with peak velocities of 90, 60 and 30°/sec respectively and parameters as shown in Table III. Five normal subjects took part in the studies which were confined to rotation in the dark with fixation upon the target light being maintained throughout the period of rotation. Fifteen stimuli were delivered
1./~o/..1,o/.
.....
2 x sd
NORMALS
PATIENTS
(J
[C
•
•
"
Fig. 11. Amplitudes of slow vertex response to 2 sec stimulus in 10 normal subjects and 12 patients with absent or reduced vestibular function. In 9 patients no response could be elicited (30 responses averaged).
in a clockwise and 15 in an anti-clockwise direction. Amplitudes in microvolts measured from the baseline are displayed individually in Fig. 9. There is a clear reduction in the amplitude of the evoked response as the peak stimulus velocity is reduced. The overall difference is significant at the level P = 0.0073. The results of applying the same analysis as before expressing phase relationships in terms of time differences revealed that as peak velocity increased from 30 to 60 to 90/sec, phase advance decreased from 342 to 259 to 230 sec respectively (level of significance: P = 0.0053). Tests on patients with absent vestibular function Adopting the same test procedures as described earlier and using the 2 sec stimulus, tests were carried out upon 12 patients with absent or grossly reduced vestibular function. Recordings from two patients are shown in Fig. 10. Nine patients had absent vestibular responses but three had either minimal caloric responses or weak nystagmic responses in total darkness to fairly powerful rotational stimuli indicating the survival of some residual vestibular function. The results in terms of the amplitude of the evoked responses, averaged for clockwise and anti-clock-
274 wise rotation, are shown in Fig. 11 together with the distribution for normal subjects. N o responses were detectable from those patients with total absence of both caloric and rotational responses. In those with some residual response to rotational stimuli, however, an evoked response was detectable but fell well below the magnitudes of the responses found in normal subjects. Nevertheless, the fact that any response at all was present in patients with apparently so little residual function is surprising and suggests that the magnitude of the evoked response may not be a direct function of peripheral vestibular sensitivity. Interestingly, in 25% of the recordings in the dark and 40% in the light an initial long latency complex was identified similar to that found in a somewhat higher proportion of normal subjects.
Discussion
An obvious and important question to which it is appropriate that we should first address ourselves is whether or not the presumed vestibular evoked potential we have described can in fact be ascribed to some other cause. Artefacts of mechanical or electrical origin would, on the face of it, appear unlikely in the light of the precautions taken. Earlier studies (Greiner et al. 1967; Spiegel et al. 1968; Bumm et al. 1970; Salamy et al. 1975; Bodo et al. 1982; Gerull et al. 1982; Hofferberth and Rothenberger 1982) have all described a vestibular evoked response which bears a marked resemblance to the initial long latency complex both in respect of latency and wave form which on some, but not all occasions, preceded the slow negative potential, the latter being the dominant feature of our own findings 3. Most of the studies, however, used stimuli of fairly abrupt onset that may well have evoked responses of non-vestibular origin. In our own studies the latency of the initial negative wave was of the order of 250 msec and the following positive wave 400 msec. However, 3 The system bandwidth was well suited to the examination of both middle and long latency components of the evoked potentials as well as the slow potentials described here.
J.D. HOOD, A. KAYAN the stimulus was of such gradual onset that we have no means of knowing at which point in time threshold activity from whatever origin was attained, sufficient to initiate a response. In the circumstances, comparison with similar long latency waves described in the literature, N1, P2, P300, and so forth would seem to be something of a pointless exercise. All that can be said is that the interval between the negative and positive waves (mean 160 msec) is much longer than the N1-P2 interval (50-70 msec). In view of the fact that similar wave forms were found in some subjects with absent vestibular function and occurred only inconstantly in normal subjects, there is a strong suspicion that they have to do with levels of arousal, vigilance and attention. There remains the question of the slow negative wave. The most telling evidence that it is vestibular in origin is, of course, our inability to detect it in patients with absent labyrinthine function. Since, however, the vestibular system interacts extensively with other sensory systems, this does not in itself rule out the possibility that it may in fact initiate a secondary response in one or more of them which manifests itself as an evoked response. Similar reservations have already been put forward by Spiegel et al. (1968) but were prompted by their own finding that ' . . . i n man either diffuse responses or responses limited to or prevalent in the pre-occipital (area 19) a n d / o r the parastriate region (area 18) were noted.' In respect of the former they suggested that these ' . . . m a y be due to impulses propagated from the second somatic sensory area that is supposed to be in the human in or around the upper bank of the sylvian fissure.' Greiner et al. (1967) likewise reported maximum vestibular evoked responses from the temporo-occipital region. In our own studies the amplitudes of the potentials from the mid-occipital, posterior temporal and parietal electrodes were found to be appreciably smaller than those recorded from the vertex. In view of the similarity of the stimuli used at least in respect of the studies of Greiner et al., the only explanation we can offer for this conflict in our findings is that it has to do with the special precautions we have taken to eliminate the influence of eye movements in the response. Both
VESTIBULAREVOKED RESPONSE Spiegel et al. (1968) and Greiner et al. (1967) rotated their subjects with closed eyes and it may be that cortical activity associated with eye movement induced by way of the vestibulo-ocular reflex made a not insignificant contribution to the evoked response. By the same token, it needs to be acknowledged that motor activity involved in the act of restraining eye movements, which would otherwise have been induced by the vestibular stimulus, may need to be taken into consideration. Although it is to be expected that if this were so then the response would follow that of the induced ocular tonus which in turn closely follows the stimulus wave form, this possibility cannot be entirely excluded. Given that the responses reported here are vestibularly evoked, any suggestions put forward as to their origin must, of necessity, be highly speculative. Latency measurements of the kind which have been so usefully applied in both visual and auditory evoked responses in the identification of the site of origin of this or that wave can have no similar relevance in respect of vestibular evoked responses because of the relatively long duration of the stimulus. However, it is perhaps worth drawing attention to certain broad parallels which can be made between the results of animal experimental studies and our own. Thus, B~ttner and Henn (1976) recorded thalamic unit activity in the alert monkey in response to a wide range of sinusoidal rotational stimuli and found the response to be progressively phase advanced as the frequency was lowered. The stimuli that we have used simulated discrete naturally occurring head movements which take the form of a raised cosine. The Fourier analysis of this contains a wide range of frequencies with most of the energy in the low frequency components so that it is somewhat meaningless to specify phase differences at any one frequency. Nevertheless it is interesting that the latencies of the main low frequency components of the responses to the 1 and 2 sec stimuli imply much the same kind of pattern as those described by Bt~ttner and Henn (1976). One finding of particular interest is the clear demonstration of visuo-vestibular interaction inasmuch as when subjects were rotated in the light
275 while fixating the target there occurred a statistically significant increase in the amplitude of the evoked response recorded in darkness. Since all eye movements were restrained this must have arisen from the images of the environment traversing the stationary retina. Convergence of visual and vestibular inputs has now been well documented at vestibular nuclei (Henn et al. 1974: Allum et al. 1976; Waespe and Henn 1977: thalamic levels - - Bt~ttner and Henn 1976; cortical levels - - Bt~ttner and Henn 1981). The main features seem to be that in isolation, phase differences are apparent in the responses to vestibular stimuli but not optokinetic while in combination algebraic summation of the two does not occur. Indeed, Waespe and Henn (1977) postulate a switching mechanism whereby the response to one stimulus may dominate the other. Strictly speaking our studies are not directly comparable to earlier investigations in which eye movement responses to optokinetic stimulation were left unrestrained. Furthermore, we have attempted unsuccessfully to record an evoked response to fullfield optokinetic stimulation with movement of the drum simulating that of the chair over a 2 sec period. In these studies the subject was required to fixate a stationary target placed close to the optokinetic drum thus suppressing all eye movements. (Earlier reports of a response under these stimulus conditions by one of us (Hood 1983) now seem likely to have been the result of an electrical artefact.) In addition we have, as yet, been unable to detect any obvious evoked response in patients with absent labyrinthine function when they were rotated in the light with eyes open. It could be that the optokinetic response in isolation is too small to be detected or that under the conditions of our experimental procedure with simultaneous visual and vestibular stimulation the former acts synergistically to enhance the latter.
Resum6
Obseruations faites sur les r@onses evoqukes ~ u n e stimulation uestibulaire naturelle Des stimulus r6p6t6s simulant les mouvements normaux de la trite ont 6t6 appliqu6s pour l'6tude
276 d e la r 6 p o n s e 6 v o q u 6 e v e s t i b u l a i r e c h e z d e s s u j e t s n o r m a u x et c h e z 12 s u j e t s p a t h o l o g i q u e s q u i o n t perdu toute fonction vestibulaire. D e s p r 6 c a u t i o n s s p 6 c i a l e s o n t 6t6 p r i s e s p o u r 6 1 i m i n e r r o u t e s les s o u r c e s p o s s i b l e s d ' a r t e f a c t , e n p a r t i c u l i e r les m o u v e m e n t s d e s y e u x o n t 6t6 r 6 d u i t s e n d e m a n d a n t a u s u j e t d e fixer le r e g a r d s u r u n e c i b l e l u m i n e u s e fix6e a u f a u t e u i l g i r a t o i r e . P o u r u n e s t i m u l a t i o n d e 2 sec la r 6 p o n s e t y p e se pr6sente comme une onde n6gative prolong6e avec u n e a m p l i t u d e m a x i m u m d ' e n v i r o n 2 3 / z V et e n r e g i s t r 6 e a u n i v e a u d u v e r t e x . C e t t e r 6 p o n s e t y p e est a b s e n t e c h e z les s u j e t s p a t h o l o g i q u e s . C h e z les s u j e t s n o r m a u x c o m m e c h e z les s u j e t s p a t h o l o g i q u e s la r 6 p o n s e est p r 6 c 6 d 6 e d ' u n e o n d e de latence prolong6e consid6r6e comme dtant d'origine non vestibulaire. Les 6preuves conduites s o i t d a n s l ' o b s c u r i t 6 t o t a l e s o i t h la l u m i + r e m o n t r e n t u n e a u g m e n t a t i o n d u p o t e n t i e l d a r t s la lumi6re indiquant ainsi une influence optokindtique. Les p h a s e s d e c h a n g e m e n t s p r o v o q u d s p a r les v a r i a t i o n s d e l ' a m p l i t u d e et d e la d u r d e d u s t i m u lus o n t dt6 e x p l o r 6 e s a u c o u r s d ' 6 t u d e s u l t 6 r i e u r e s . C e l l e s ci o n t r6v616 c e r t a i n e s s i m i l a r i t 6 s a v e c les rdsultats d'exp6riences r6centes chez l'animal. Our grateful thanks are due to Mr. Y.L Chua, Mr. E. Trinder, Mr. A. Beechey and Miss D. Jackaman for technical assistance and to Dr. J. Stephens for the translation of the abstract.
References Allure. J.H., Graf, J.W., Dichgans, J. and Schmidt, C.L Visual-vestibular interactions in the vestibular nuclei of the goldfish. Exp. Brain Res., 1976, 26: 463-485. Bodo, Gy., ROzsa, L. and Antal, P. Scalp electrical response evoked by acceleration in young healthy man. In: L. Surjan and Gy. BodO (Eds.), Borderline Problems in Otorhinolaryngology. Excerpta Medica, Amsterdam, 1982: 389-392.
J.D. HOOD, A. KAYAN Bumm, P., Johannsen, H.S., Spreng, M. und Wiegand, H.P. Zur Registrierung langsamer Rindenpotentiale bei rotatorischer Reizung des Menschen. Arztl. Forsch., 1970, 24: 59-62. Btittner, U. and Henn, V. Thalamic unit activity in the alert monkey during natural vestibular stimulation. Brain Res., 1976, 103:127 132. Biattner, U. and Henn, V. Circularvection: psychophysics and single-unit recordings in the monkey. Ann. N.Y. Acad. Sci., 1981, 374: 274-283. Donchin, E., Callaway, E., Cooper, R., Desmedt, J.E., Goff, W.R., Hillyard, S.A. and Sutton, S. Publication criteria for studies of evoked potentials (EP) in man. Report of a committee. In: J.E. Desmedt (Ed.), Attention, Voluntary Contraction and Event-Related.Cerebral Potentials. Prog. Clin. Neurophysiol., Vol. 1. Karger, Basel, 1977:1 11. Gerull, G., Gieseu, M., Keck, W. und Mrowinski, D. Rotatorisch evozierte Hirnrinden-Potentiale beim Menschen. Biomed. Technik, 1982, 26:262 266. Greiner, G.F.. Collard, M., Conraux, C., Picart. P. et Rohmer, F. Reche/che des potentiels 6voquds d'origine vestibulaire chez l'homme. Acta oto-laryng. (Stockh.), 1967, 63:320 329. Henn. V., Young, I.R. and Finley, C. Vestibular nucleus units in alert monkeys are also influenced by moving visual fields. Brain Res., 1974, 71:144 149. Hillyard, S.A. and Galambos, R. Eye movement artefact in the CNV. Electroenceph. clin. Neurophysiol., 1970, 28: 173-182. Hofferberth, B. and Rothenberger, A. Event related potentials to rotatory stimulation preliminary results in adults and in children. In: A. Rothenberger (Eds.), Event Related Potentials in Children. Elsevier Biomedical Press, Amsterdam, 1982: 463-470. Hood. J.D. Vestibular and optokinetic evoked potentials. Acta oto-laryng. (Stockh.), 1983, 95:589 593. Salamy, J., Potvin, A., Jones, K. and Landreth, J. Cortical evoked responses to labyrinthine stinmlation in man. P~ychophysiology, 1975, 12:55 61. Spiegel, E.A., Szekely, E.G. and Moffet, R. Cortical responses to rotation. 1. Responses recorded after cessation of rotation. Acta oto-laryng. (Stockh.), 1968, 66:81 88. Waespe, W. and Henn, V. Neuronal activity in the vestibular nuclei of the alert monkey during vestibular and optokinetic stimulation. Exp. Brain Res., 1977, 37:337-347.