close

Вход

Забыли?

вход по аккаунту

?

Binaural interaction in brainstem potentials of human subjects.

код для вставкиСкачать
Binaural lnteractlon in Brainstem
Potentials of Human Subjects
R o b e r t A . Levine, MD
Binaural interaction i n the short-latency averaged auditory evoked potentials (AEPs) can be assessed from the
binaural difference waveform (BD). The B D is derived b y c o m p u t i n g the difference between the AEP evoked by
simultaneous clicks f r o m b o t h earphones and the s u m o f two other AEPs: o n e evoked by clicks from the r i g h t
earphone alone a n d the other evoked b y clicks f r o m the left earphone alone. Once the contributions of acoustic
cross talk a n d the middle ear reflex are eliminated, the B D can b e considered to represent neural binaural interaction. T h i s interaction begins after wave I11 and has its first major p e a k during the downslope of wave V. The
relative amplitude of this peak to wave V is constant over a wide r a n g e of click rates and levels. The B D may have
more localizing value than the AEPs alone since the B D probably represents a subpopulation of generators of the
AEPs, which s h o w binaural interaction at the single cell level.
Levine RA: Binaural interaction in brainstem potentials of human subjects.
Ann Neurol 9:384-393, 1981
Recording short-latency (<10 msec) auditory e v o k e d
potentials (AEPs) from electrodes placed on t h e
scalp of h u m a n s [14] has e n h a n c e d t h e ability to
d e t e c t disorders involving t h e brainstem [22]. Animal studies have s h o w n that with some lesions,
changes can occur i n AEP c o m p o n e n t s that exhibit
binaural interaction [ 5 , 61. Presumably t h e s e cornponents reflect activity i n structures that are binaurally
activated. T h u s , i t appears that binaural interaction i n
human AEPs n e e d s a more detailed description than
is presently available [2, 3 , 9, 2 3 , 251.
Methods
Twenty-five adults (9 men and 16 women ranging in age
from 20 to 40 years) with normal audiograms and n o
known neurological disease served as normal subjects. Six
others, each of whom had one deaf ear and one ear with
normal hearing, were also studied. T h e deaf ear did not
respond by ordinary audiometric measurements. The deafness was idiopathic in 2 subjects and in the others was attributable to meningitis, mumps, surgical removal of an
acoustic neuroma, or labyrinthectomy.
Recordings were made with the subjects lying supine in a
darkened, sound-treated room [24].Stimuli were rarefaction clicks produced by delivering 30-psec electrical pulses
to shielded Sennheiser HD424 earphones in a standard
headband with a chinstrap. Responses were recorded differentially between surface electrodes at the vertex and the
nape of the neck, just below the hairline. T h e ground was a
forehead electrode. The potentials were amplified and
bandpass-filtered (from 1 to 3,000 Hz, -3 d b cutoffs),
From the Neurology Service, Massachusetts General Hospital,
Boston, MA 02114, and the Department of Otolaryngology,
Eaton-Peabody Laboratory of Auditory Physiology, Massachusetts
Eye and Ear Infirmary, Boston, MA 02114.
384
sampled at 80 psec intervals, and averaged over a 20 msec
interval beginning 8 msec before the click (to establish a
baseline level of activity). To limit the effects of muscle
activity, a circuit to reject large deflections was employed.
For each stimulus condition the averaged response to 2,048
clicks was obtained and then digitally bandpass filtered (-3
d b cutoffs of 10 and 1,600 Hz; 6 dh per octave slopes)
using the formula:
where xi are the values of the points of the original
waveform, yi are the values of the points of the resultant
filtered waveform, and y,, is set to zero.
Three stimulus conditions were repeated at least twice
for each combination of click level and rate. Averaged responses were obtained to 2,048 clicks delivered to the right
earphone alone (monotic right), to the left earphone alone
(monotic left), and to both earphones simultaneously (diotic). T h e three final average responses were based o n at
least two runs of 2,048 trials each, or 4,096 trials in all.
Click levels were expressed in decibels relative to an average of the click thresholds for a group of 30 normal subjects (thresholds < 20 db HL at the six standard audiometric frequencies). Typically in a session, 10 clicks per second
were used at levels between 8 and 68 db. With 5 selected
normal subjects a second session was held in which five
click rates (10, 20, 30, 50, and 70isec) were used at 38 dh.
Positive peaks in the response waveforms were labeled
according to Jewett’s 1121 convention (see Fig l), and
negative peaks were numbered by the preceding positive
Received June 5 , 1980, and in revised form Aug 4. Accepted for
publication Sept 10. 1980.
Address reprint requests to D r Levine, Eaton-Peabody Laboratory,
Massachusetts Eye and Ear Infirmary, 243 Charles St, Boston, MA
02114.
0364-5 134/81/040384-10$01.25 @ 1981 by the American Neurological Association
peak. Latencies were measured from the time the electrical
pulse reached the earphone.
For 8 normal subjects a Grason-Stadler Middle-Ear
Analyzer (Model 1723, 220-€Iz probe) measured the
acoustic impedance of one ear while clicks were delivered
to the other ear in the same manner as for the AEPs. The
threshold for eliciting an impedance change was determined to within 2 d b for click rates of 10, 20, 30, 50, 7 0 ,
and 100 per second. The threshold was the lowest click
level to deflect the meter by 0.01 ml from its baseline.
Results
For six click levels the AEPs to monotic right, monotic left, and diotic clicks are shown in Figure 1. At
each level the diotic potentials are larger than either
of the monotic potentials, with the components of
the AEPs generally increasing in amplitude and decreasing in latency as level is increased.
Binaural interaction in the AEPs was assessed by
calculating the binaural difference waveform (BD)
(Fig 2 ) by subtracting the diotic potential,
from the sum of the monotic potentials, P(R) P,,). If
no interaction occurs between the inputs from the
two ears, then the diotic potential should equal the
sum of the monotic potentials, and the B D will be
flat. With binaural interaction present the diotic potential will differ from the sum of the monotics, and a
+
F i g I . Monotic and diotic AEPs for six click lwels. Click rate
i.c 10 per second. The responses were recorded between electrodes
on the uertex and on the nape ofthe neck; vertex positivity i s
plotted ujward. Zero time, marked by a short vertical bar,
represents the time the electric pulse occuu at the earphone. The
positive peaks are labeled with Roman numerals according to
Jezuett's convention [12].
RIGHT
EARPHONE
dB
STIMULATED
\o
nonzero BD will result. Figure 3 shows the BDs for
six click levels for one subject and five click rates for
another subject. The first positive peak in the B D
will be referred to as p, the second positive peak as 6,
and the first negative peak as a.
For waves I, 11, and 111, the sums of monotic traces
are essentially identical to the diotic traces so that the
B D is initially flat. After III( -), however, a nonzero
B D appears. In Figure 4A, the time of the earliest
appreciable nonzero binaural difference is plotted
against the time of III( -) of the P(R)+ P,,,) waveform
for all subjects and stimuli. The first large positive
peak in the BD (p in Fig 3A) occurs consistently
during wave V. Plotting its latency against that of
v(+) and V( -) of P(R) + P,,,,, Figure 4 B and c shows
that for all stimuli, p occurs between V(+) and V( -),
i.e., on the descending limb of wave V (see Fig 3).
This finding is consistent with the fact that the latency
of V( +) of P(R)+ P,,,)is later than V( +) of P,R+I,l
(Fig
4D).
p amplitude as a fraction of wave V amplitude (of
P,,, + Po,) has a mean ratio of about 0.2 that is constant for 10 per second clicks over a wide range of
levels (Table 1).For higher click rates this ratio tends
to become smaller (Table 2). This effect, which may
be related to activation of the middle ear reflex, is
discussed later.
Provided the clicks were at a sufficiently high level,
p was identified in all subjects (Table 1). Sometimes a
smaller negative peak, a , preceded p (e.g., Fig 2 and
the 58 and 68 d b traces of Fig 3A). a occurs when the
sum of the monotic AEPs is more negative than the
diotic AEP. It generally appears after wave IV(+) and
tends to coincide in time with wave IV(-). The
LEFT
EARPHONE
BOTH
STIMULATED
STIMULATED
EARPHONES
\ , {
/I
68
58
48
38
4
4
28
18
4
-
0
4
8
l2msec
0
4
8
12msec
SUBJECT 66
7 . 7 7 1
0
4
0 12msec
Levine: Binaural Interaction in AEPs
385
BOTH
EARPHONES
RIGHT EARPHONE
STIMULATED
TR,
COMPUTED
I
[%I+
LEFT
ED
S,J- [ % + L J
J
EARPHONE
... .
I
I
’
0
I
4
I
124
SUBJECT
I
‘
0.6 p V
I
12msec
8
F i g 2. Deriving the B D . Plotted on the left side of the figure
are the diotic (P(R+I.J),
monotic right (P(R,),and monotic left
( P(I,l)waveforms for a click rate of 10 per second and level of 28
db. The dotted waveform was obtained by adding P,,, and P,,,.
On the right, P(R+I.J
and P,,, + P,,, are superimposed. Below
them is the BD, which was calculated by subtracting the
former from the latter.
COMPUTED
LEVEL
CLICK
(68
d6)
38
BD
2
-
COMPUTED BD
(per sec)
10
Md>
50
28
18
0.6
.VI
&
A
SUBJECT 66
TTlrlrl
0 4
8 12msec
A
70
0.6 I.rv
1
SUBJECT 152
0
4
8
12msec
B
F i g 3 . (A)BDs for six click levels computed from the
waveforms i n Figure 1. Clrck rate t s 10 per second. Arrowheads mark the time of occurrence of peak V (+) i n the P,,,
iP,,,, waveform. ( B ) BDsforfive click rates. Click level is 3 8
db. The steep dope of the prestrmulus portron of the 20 per
second waveform i s due t o some of the longer- latency potentials
evoked by the clicks.
386 Annals of Neurology
Vol 9
No 4 April 1981
0
V
In
v)
E
a1
/
/
/
'.
ZlD
/
I
,
,
n = 121
0
-
0
ml"
0
[x-y)=-109
,,
,
S D =030
S E =O027
4
5
6
PEAK I D M OF
7
XlD
a
w
,
5
Bmsec
\TRl+F;Ll]
A
,
PEAK
SD.019
S E =0017
6
8
7
9msec
[TR)+yL)L,]
Y(+) OF
B
,
/
,
/
I
/
/'
S D. = 0.26
S E. = 0.023
I,/', , , , ,
5
PEAK
L
6
7
Y(-) OF
8
5
9msec
+R)]
6
7
PEAK Y(+) OF
D
F i g 4. Four scatter plots. Each point represents a pair of latency measurements made upon a single group of waveforms
similar t o those of Figure 2. All 149 such groups that comprise
this study are represented i n each plot, except when the component being measured could not be recognized. Indicated on each
plot is the total number of runs plotted (n), the mean (-1,
standard deviation (S.D.),and standard error of the mean
(S.E.) of the absci.rsa value minus the ordinate value of each
point. ( A )Latency of I l l ( - ) of P,,, + P,,, versus onset of
binaural interaction in B D . I l l ( -) is the negative peak that
follows I l l ( + ) . Onset of binaural interaction i s the time the
first significant peak of the B D waveform seems t o begin. lB)
Latency ofV(+J o f P,,, t P(,,,versus latency of /3 peak of BD.
(CJ Latency of V( -i of P,,,, + P,,,, versus latency of p peak of
B D . V (-) is the negative peak or injection that follows V( +i.
l D ) Latency of Vt +i of P,,, t P,,,, tjersus latency of V ( + i of
P<R+I.).
smaller amplitude of (y makes it more difficult to detect than p.
The B D is generally reproducible within a recording session and from session to session. Each
waveform in Figure 2, for example, represents the
average of six (2,048 click) segments. Five of the six
BDs were similar; only the one with the highest prestimulus noise level showed any appreciable variation.
For 4 other subjects, the two BDs obtained up to 12
months apart for identical stimuli (10/sec, 38 db
clicks) were invariably similar. On one occasion, for a
different subject, no clear binaural interaction could
Table I . Percentages of Subjects i n Whom a and p Peaks
Were ldenti5ed and Relative Amplitude of p for 10 per
Second Clicks at Various Levels
Click
Level (db)
68(N
58(N
=
=
48(N =
38(N =
28(N=
18(N =
8(N =
22)
13)
16)
15)
15)
11)
10)
BD Peaks
Identified
(%)
a
36
38
25
33
13
9
0
P
100
100
94
80
73
64
10
PlVC + )
Amplitude Ratio"
Mean"
SD
SEM
0.19
0.19
0.20
0.19
0.19
0.19
0.17
0.08
0.09
0.08
0.07
0.06
0.06
...
0.02
0.02
0.02
0.02
0.02
0.02
...
" p amplitude is measured from the point of the BD waveform immediately before interaction occurs. V( +) amplitude is measured
in the P,,, + P,,.) waveform from wave HI(-).
t'For the runs in which p could be identified.
Numbers in parentheses are total number of subjects. B D =
binaural difference waveform.
be found for a run of 6,144 clicks (lolsec, 68 db), but
when the same run was repeated a second time in the
same session, a typical BD resulted. The cause of the
different results is unknown but may have been due
to breathing noises (causing masking) or shifts in the
earphone position (thereby altering the sound
reaching the tympanic membrane).
To determine whether the BD is entirely due to
Levine: Binaural Interaction in AEPs
387
Table 2. Relative p Amplitude for 38 db CLicks
at Various Rates
Click Rate
( p e r sec)
10(N =
20(N =
30(N =
50(N =
70(N =
5) .
5)
5)
5)
5)
ON(+)
Table 3. Mean Thresholds for Acoustic Impedance Change
Elicited by Monotic Contralateral ClickJ f a r 8 Normal Subjects
Amplitude Ratio"
Mean
SD
SEM
0.16
0.19
0.16
0.12
0.12
0.05
0.07
0.05
0.03
0.05
0.02
0.03
0.02
0.01
0.02
" p amplitude I S measured from the point of the BD waveform immediately before interaction occurs. V( +) amplitude is measured
in the P,,,
P,,,) waveform from the III(-) peak.
+
Numbers in parentheses are total number of subjects. BD =
binaural difference waveform.
neural interactions, we studied two possible extraneous factors: the middle ear reflex (MER) and acoustic
cross talk (ACT). Because middle ear muscle contractions reduce sound transmission through the
middle ear [la, 181 and are greater for diotic than
monotic sounds [ 161, during diotic runs each cochlea
could receive clicks that are at a lower level than
during the monotic runs. Since the AEPs vary with
click level (see Fig l), this level difference between
monotic and diotic runs could result in a nonzero
BD.
At least three lines of reasoning suggest that the
influence of the MER upon the BD is negligible.
First, the MER attenuates predominantly low-frequency components of sounds [ l a , 181. The click
level at the cochlea is probably at most 5 d b lower for
diotic than for monotic stimulation [16]. Moreover,
since this attenuation occurs before the clicks reach
the cochlea, a similar effect might be expected on all
the waves of the AEPs. Yet there is n o consistent BD
for waves I, 11, and 111. Second, if the MER attenuates the diotic more than the monotic clicks, the
latency of V(+) should be longer for the diotic than
for the monotic AEPs, since latencies increase for
lower click levels (see Fig 1). However, Figure 4 D
shows just the opposite result.
To determine the extent to which the MER was
active during our recordings, we measured in 8 subjects the threshold for an acoustic impedance change
in one ear while monotic clicks were being presented
to the other ear (Table 3). With increasing click rate
the threshold generally decreased. The threshold for
the ear ipsilateral to the clicks is about 2 to 16 d b
lower than for contralateral clicks, and another 3 d b
lower for diotic clicks [16]. For clicks 19 d b below
our measured thresholds, then, the MER should not
cause any sound attenuation 1181. In runs with such
stimuli, clear a and p peaks occurred consistently
(e.g., Figs 2 and 3 and Tables 1 and 2). Thus, while
388 Annals of Neurology
Vol 9 No 4 April 1981
Click Rate
( p e r sec)
10
20
30
50
70
100
Impedance Change (db)
Mean
SD
78.4"
64.6''
55.1
46.4
39.4
41.1
7.4
8.9
11.1
8.9
9.7
8.1
"In cases for which no change could be detected at the maximum
click level of 84 db (4 at lO/sec, 1 at 20/sec), a threshold of 84 db
was used for calculations.
the influence of the MER cannot be absolutely excluded at high levels and rates, the presence of nonzero B D for stimuli where the MER is almost certainly not active strongly suggests that the a and p
peaks are not attributable to activity of the middle ear
muscles.
T h e second nonneural factor analyzed was ACT. In
4 subjects, as clicks were being presented by one of
the earphones, the sound pressure was measured at
the entrance to both external auditory canals. (Figure
5 shows how the waveform recorded contralaterally
is an attenuated, delayed, and modified version of the
click waveform recorded ipsilaterally.) ACT was also
demonstrated in monaurally deaf subjects by
measuring behavioral thresholds to monotic clicks.
The mean threshold difference between the two ears
was 49 db (SD = 2 db). At click levels below 43 d b
(see Figs 2 and 3 ) , then, the CY and p peaks of the B D
cannot be due to ACT. However, a third peak, 6 (see
Fig 3A), appears only at high click levels and might
be due to ACT.
The effect of ACT upon the AEPs was assessed in
3 normal subjects by simulating from one earphone
the stimuli received by the underlying ear during ipsilateral, contralateral, and diotic clicks (Fig 6A). s,,,
(delivered at time zero and 68 db) was the monotic
ipsilateral click. Sfz,, a click delayed by 1 msec and
attenuated to 2 8 db, simulated the 68 db monotic
click from the other earphone. P(,) and P,,, are the
corresponding AEPs, while P,,,,, is the waveform
evoked by the simulated diotic condition, S,,, The
simulated binaural difference waveform, [P,,)+ P,,)]
- [P(l,,,], has a prominent peak with a latency like
that of 6 of the BD. Note that P(,+,, is virtually identical to P(l,. Apparently, in the simulated diotic condition, the "cross talk" click, &), does not elicit its
normal response. Presumably this effect is due to refractoriness in the system, such as has been described
by Rosenzweig and Rosenblith [ 191.
,,.
SPECTRAL
SOUND PRESSURE
WAVEFORM
DENSITY
I
1:
I
I
x
,,$
0
j
RECORDING
MICROPHONE
CONTRALATERAL
10 I dynes/cm2
II
0
-40
I
2
3
4
Smsec
-60'
1
1
I
02
1
I I1l'll
05
F i g 5 . Acoustic responses, as measured under the earphones
near the entrance t o the external auditory meatus of each ear,
for 10 per second monotic clicks at 78 db. Tracings from the
side with the active earphone are dotted (top row); tracings
from the opposite side are solid (bottom row). Recording was
done with a '14-inch condenser microphone (Bruel and Kjaer,
type 41361, which was flat t o within I db from 50 t o 15,000
Hz. The tracings on the left show averaged sound pressure
iuaQeforms.Note the difference i n the vertical scales for measurements at the two ears. The magnitudes of the power spectra
computed from the two waveforms on the left (using a Fast
Fourier Transform) are plotted on a single set of axes on the
right.
Monaurally deaf subjects provide an opportunity
for more direct examination of ACT effects, as seen
in Figure GB. Monotic clicks on the deaf side evoked
a small potential, P(L),that must be due to ACT. O n
the other hand, during the diotic run these same
clicks apparently do not evoke any potentials, because the diotic AEP is virtually identical to the
monotic right AEP (note Po+,, - P(R)).Presumably
the normal ear is refractory to the cross talk clicks
during diotic stimulation, so a peak like 6 appears in
the BD. Thus, in normal subjects (Fig 7) 6 can be
interpreted as follows. For high-level monotic clicks,
responses attributable to both ears are present in the
AEPs: the ipsilateral ear responds to the direct click
and the opposite ear to the cross talk click. For diotic
clicks, however, only the responses attributable to
the direct clicks are present because neither ear responds to the delayed and attenuated cross talk
click. Consequently, at high intensities the B D will
have a component that results from ACT rather than
from neural interaction. This point is very important
in any study of binaural interaction. It was not ap-
10
I
20
I
I
I
1 1 1 ' 1
50
10
I
20kHz
preciated in at least one preliminary description [ 3 ]
of binaural interaction in human subjects.
By reducing the air-conducted ACT during monotic stimulation, its contribution to the BD can be assessed. In Figure 8 the contralateral ear was occluded
for one series and unoccluded for another series of
monotic runs. For the 68 db monotic clicks, the open
arrows in Figure 8 point to the portion of the AEPs
that decreased with occlusion of the contralateral ear.
This effect may account for some of the variability in
this portion of the AEPs [21]. Comparison of the
BDs shows that attenuating the air-conducted ACT
markedly reduces 6, albeit incompletely.
Bone-conducted ACT might also be contributing
to the BD. T o estimate the stimulus levels at which
bone conduction might be important, in 6 monaurally deaf subjects we determined the behavioral
thresholds for monotic clicks on each side before and
after occlusion of the normal ear. If occlusion may be
considered to attenuate mainly air-conducted sounds,
the threshold for bone-conducted cross talk clicks
must then be at least as high as the threshold for
monotic clicks on the deaf side with the normal ear
occluded. The difference between this conservative
estimate of threshold for bone-conducted clicks and
the threshold of the normal unoccluded ear was
taken as a minimum estimate of the interaural attenuation for bone-conducted clicks. For these subjects
the attenuation ranged from 61 to 71 db, with a
mean of 65 db.
For the subject of Figure 8, click thresholds were
-6 d b for the right and -4 d b for the left ear. At 53
db (right column in the figure) the bone-conducted
clicks would be below these thresholds and therefore
should not be contributing to the AEPs. Not only are
the BDs similar for the occluded and unoccluded
Levine: Binaural Interaction in AEPs
389
SIMULATED ACOUSTIC CROSSTALK
S,,)T O
RIGHT E A R
( 6 8 dE)
SUBJECT 141
(DEAF LEFT EAR)
72,
S
TO RIGHT EAR
( 2 8 dB)
S,,,,,TO
RIGHT E A R
1,
%+L)
- %i
.........
. . . . . . . ............................
.
SUBJECT I51
A
0
4
conditions at 53 db, but binaural interaction is present well beyond the time of /3. Thus, some neural
binaural interaction is present in the 6 peak of the
BD, even though air-conducted ACT may account
for a large part of it at high levels. Bone-conducted
ACT probably does not contribute to the BD for the
lower-level stimuli of this study, but at higher levels
this possibility is not excluded.
Discussion
This study shows that a nonzero binaural difference
waveform need not imply neural binaural interaction
unless certain experimental factors are taken into account. One of these factors, acoustic cross talk, can
give a nonzero BD at high click levels. Another, the
middle-ear reflex, did not appear to contribute
significantly to the BD for the click levels and rates
used here but could be a factor under other conditions.
Some practical advice may be offered about the
combinations of click rate and level for which the BD
might be expected to be free of these nonneural factors. The exact levels, of course, will vary for different headsets. The maximum level to avoid ACT is 46
d b above the threshold of the opposite ear, since the
minimum interaural attenuation for 10 per second
clicks in the monaurally deaf subjects was 46 d b
(mean, 49 db). A level of about 5 1 d b is probably still
safe because the AEPs usually cannot be detected
until they are at least 5 d b above threshold. T h e MER
threshold diminishes as click rate increases. Clicks 19
390 Annals of Neurology Vol 9 No 4 April 1981
8
12rnsec
0
4
8
12msec
B
F i g 6. How A C T may have different eHerts during monotic
and diotic stimulation. (A) Simulated A C T for one ear of a
normal subject: AEPs t o weak and more intense clicks il Olsec)
t o the same earphone. P,,, is evoked by the intense stimulus S,,,
alone, presented at 68 db at zero time. P,,, is evoked by the
weak click S,,,, presented alone at 28 db and delayed 1 msec
from zero time. Pi,+z, is the poteniial evoked by both stimuli,
S, ,+2,, presented together. Calculated d;fference waveforms:
[Pi,, + PI,,] - [P,,,,,] is analogous t o the B D and shows a
waveform with a peak similar to 6 of the B D . P,,, is similar t o
P,,,,, as shown by the P(,+,, - PI,,wavefirm. ( B i B D i n a
subjert deaf in the IeJt ear; 1Olsec r1irk.r at 68 db. The subject
was a neurologically intact 26-year-old woman with no uudiometrically detectable hearing i n the left ear due to meningitis at age 2 months. The B D, ulhich i s z)irtually identical
t o P,,,,, must be due t o A C T . The dotted waveform, PtR+,.,
- P,,,, shouls that PI,, is similar tn PI,,+,,,.
d b below the lower range of impedance thresholds to
contralateral monotic clicks (see Table 3) is a conservative estimate of the MER thresholds for diotic
clicks [16]. For stimuli below these levels the MER
should not be active. As suggested earlier in this
paper, even if active, the MER in general might have
only a negligible effect on the BD.
For some individuals (6% at 48 db) p was not seen
at click levels below 51 db, so higher click levels
where ACT may be a factor had to be used to demonstrate the presence of p (see Table 1). One way to
reduce the ACT is to occlude the ear contralateral to
the active earphone (see Fig 8). However, unless
STIMULATED
EARPHONE
EVENTS
ATTRIBUTED
TO LEFT EAR
EVENTS
ATTRIBUTED
TO RIGHT EAR
EVOKED
RESPONSE
I
RIGHT
LEFT
10'
BOTH
COMPUTED 8D
--
A[TR)'
['(R)'
i-T--rv
0
4
12msec
8
0
4
12msec
8
F i g 7 . Model of B D due t o ACT. A monotic stimulus, S,,, or
S,,,,, produces one click at the ipsilateral ear and a delayed attenuated click at the opposite ear (due to A C T ) . From the ipsilateral ear the monotic stimulus will evoke a large earlierpotential and, from the contralateral ear, a small later potential.
The electrodes will record the sum of these two potentials as
diagramed in the Evoked Response column. Each ear will re-
68 dB
-NO OCCLUSION
CONTRALATERAL EAR OCCLUDED
0
4
8
?L,]
- [ ?R+LI]
'(LI]
[ '(RILI]
12msec
ceive an immediate unattenuated click and a delayed attenuated click to the diotic stimulus S(R+L,;
but each ear i J refractory t o the delayed (cross talk) click, so only the direct click
evokes a potential. The B D will have a positive peak due t o the
different efhcts of ACT upon the potentials during diotic and
monotic stimulation.
53 dB
nRb
-,Ju
n
LEFT EARPHONE
STIMULATED
1
F;L,
%
t
.
o
.
4
.
8
i
12msec
I
SUBJECT 159
i l I I I
6 4 6
F i g 8. Effect of occlusion of the contralateralear during monotic stimulation for two click levels. For one series of monotic
runs (dotted waveforms) the contralateral ear was occluded
with a sponge earplug that raised the ear's thresholdfor clicks
Ikrnsec
by 32 t o 42 db. Open arrows point t o portions of the monotic
waveforms that change with occlusion. Click rate is 10 per
second.
Levine: Binaural Interaction in AEPs
391
great care is taken in repositioning the earphones, a
substantial change in the actual click reaching the
tympanic membrane may result. Insert earphones
may increase interaural attenuation; however, those
presently available deliver less output at high frequencies than the earphones we use. Still another approach is to improve the earphone cushion without
increasing discomfort. Masking of the contralateral
ear is widely used for reducing ACT, but the masker
might cause central masking of the AEPs, and it may
evoke the MER.
Obtaining the B D takes longer than obtaining the
two monotic AEPs. Besides the extra time to record
the diotic AEPs, the BD's smaller peaks (see Tables 1
and 2) and larger noise level make longer runs necessary to improve its signal-to-noise ratio. One technique for improving the peak size of the B D is to
record differentially between the vertex and nape.
While recording concurrently between vertex and
earlobe throughout this entire study, we found that
wave V and the B D peaks were consistently larger in
the vertex-to-nape recordings.
Once ACT and the MER have been eliminated
from the BD, the remaining binaural interaction
(presumably neuronal) occurs after wave I11 and has a
first major peak during the downslope of wave V, the
amplitude of which, relative to that of wave V, is
constant for a wide range of click parameters.
The part of the B D due to neuronal binaural interaction can be thought of as representing a specific
subpopulation of all the neuronal elements generating the AEPs. This subpopulation would presumably
consist of cells that show binaural interaction at the
single-unit level. Such cells have been found within
most of the auditory nuclei of the brainstem [4, 7,
151. Few of these animal studies have used clicks and
none presents the far-field AEPs, so the temporal
characteristics of the units can be neither compared
with the AEPs nor related to the present results.
Binaural interaction of AEPs has been reported for
cat [5, 9-11, 13, 201 and guinea pig [ 3 , 61, and it
appears to be similar for the two species. Because the
relative numbers of cell types within the auditory nuclei vary considerably among species [17], a comparative approach may prove useful in identifying the
cells that contribute to any particular component of
the AEPs, including the BD.
One psychophysical test of binaural hearing has already been shown to be related to the monotic AEPs
[ 8 ] . Studying the relationship between the B D and
such tests may clarify which cell populations are important for different aspects of binaural hearing.
Because the BD is smaller than the AEPs, BD abnormalities could go unnoticed when only AEPs are
examined. Likewise, the B D could be normal when
the AEPs are abnormal. Since the BD, unlike AEPs,
392 Annals of Neurology Vol 9 No 4 April 1981
is generated by only binaurally activated cells, the
BD may have more localizing value than the AEPs
alone.
Add endurn
Since this paper was submitted, two reports on human
binaural interaction have appeared. The results of one [3a]
are consistent with those reported here. The other [ I ] did
not find definite neural binaural interaction, possibly because recordings were made for relatively small numbers of
stimulus presentations between vertex and ear electrodes.
Supported in part by US Public Health Service Grants 5 PO1
NS13126-03 and 5S07 RR05485.
Thanks are due Dr N. Y . S. Kiang for his suggestions and criticisms during the preparation of this paper; Drs A Thornton and K.
Chiappa for reviewing the manuscript; and L. Miller, D. Leming,
Dr H. Hosford-Dunn, and Dr W. Rabinowitz, who assisted in
various phases of this study.
References
1. Ainslie PJ, Boston JR: Comparison of brain stem auditory
evoked potentials for monaural and binaural stimuli. Electroencephalogr Clin Neurophysiol 49:291-302, 1980
la. Borg E: A quantitative study of the effect of the acoustic
stapedius reflex on sound transmission through the middle-ear
of man. Acta Otolaryngol (Stockh) 66:461-472, 1968
2. Davis H: Principles of electric response audiometry. Ann
Otoi Rhino1 Laryngol 85:suppl 28, 1976
3. Dobie RA, Berlin CI: Binaural interaction in brainstemevoked responses. Arch Otolaryngol 105:391-398, 1979
3a. Dobie RA, Norton SJ: Binaural interaction in human auditory evoked potentials. Electroencephalogr Clin Neurophysiol 49:303-313, 1980
4 Erulkar SD: Physiological studies of the inferior colliculus and
medial geniculate complex. In Keidel WD, Neff W D (eds):
Handbook of Sensory Physiology. Berlin, Springer, 1975, vol
5 , part 2, pp 145-198
5. Fullerton BC, Hosford HL: Effects of midline brainstem lesions o n the short-latency auditory evoked responses. SOC
Neurosci Abstr 5:20, 1979
6. Gardi J N , Berlin CI: A preliminary report on the origin(s) of
the binaural-interaction component of the BSER in the guinea
pig. SOCNeurosci Abstr 5:20, 1979
7 Goldberg JM: Physiological studies of auditory nuclei of the
pons. In Keidel WD, Neff W D (eds): Handbook of Sensory
Physiology. Berlin, Springer, 1975, vol5, part 2 , p p 109-144
8 Hausler R, Levine RA: Brainstem auditory evoked potentials
are related to interaural time discrimination in patients with
multiple sclerosis. Brain Res 191:589-594, 1980
9 Hosford HL, Fullerton BC, Levine RA: Binaural interaction
in human and cat brain stem evoked responses. J Acoust Soc
Am 65:S86, 1979
10 Huang C-M, Buchwald JS: Factors that affect the amplitudes
and latencies of the vertex short latency acoustic responses in
the cat. Electroencephalogr Clin Neurophysiol 44: 179- 186,
1978
11 Jewett DL Volume conducted potentials in response to auditory stimuli as detected by averaging in the cat. Electroencephalogr Clin Neurophysiol 28:609-618, 1970
12 Jewett DL, Williston JS: Auditory-evoked far fields averaged
from the scalp of humans. Brain 94:681-696. 1971
13 Kemp EH, Robinson EH: Electric responses of the brainstem
to bilateral auditory stimulation. Am J Physiol 120:316-322,
14.
15.
16.
17.
18.
1937
Kiang NYS: The use of computers in studies of auditory neurophysiology. Trans Am Acad Ophthalmol Otolaryngol
65:735-747, 1961
Mast TE: Binaural interaction and contralateral inhibition in
the dorsal cochlear nucleus of the chinchilla. J Neurophysiol
33:108-115, 1970
Meller A: Acoustic reflex in man. J Acoust Soc Am
34:1524-1534, 1962
Moore JK, Moore RY: A comparative study of the superior
olivary complex in the primate brain. Folia Primatol (Basel)
16135-51, 1971
Rabinowitz WM: Acoustic-reflex effects on the input admittance and transfer characteristics of the human middle-ear.
PhD thesis, Massachusetts Institute of Technology, Cambridge, MA, 1977
19. Rosentweig MR, Rosenblith WA: Responses to successive
auditory stimuli at the cochlea and at the auditory cortex.
Psycho1 Monogr 67:l-26, 1953
20. Rosenzweig MR, Sutton D: Binaural interaction in lateral
lemniscus of cat. J Neurophysiol 21:17-23, 1958
2 1. Rowe MJ 111: Normal variability of the brain-stem auditory
evoked response in young and old adult subjects. Electroencephalogr Clin Neurophysiol 44:459-470, 1978
22. Starr A: Sensory evoked potentials in clinical disorders of the
nervous system. Annu Rev Neurosci 1:103-127, 1978
23. Van Olfen AF, Rodenburg M, Verwey C : Distribution of
brainstem responses to acoutic stimuli over the human scalp.
Audiology 17:511-518, 1978
24. Ver IL, Brown RM, S a n g NYS: Low-noise chambers for auditory research. J Acoust SOCAm 58:392-398, 1975
25. Wrege KS, Starr A: Binaural interaction in auditory brainstem
potentials. Soc Neurosci Abstr 5:208, 1979
Levine: Binaural Interaction in AEPs
393
Документ
Категория
Без категории
Просмотров
2
Размер файла
755 Кб
Теги
potential, binaural, interactiv, brainstem, human, subjects
1/--страниц
Пожаловаться на содержимое документа