PreprintPDF Available

Evidence for a system in the auditory periphery that may contribute to linking sounds and images in space

Authors:
Preprints and early-stage research may not have been peer reviewed yet.

Abstract and Figures

Eye movements alter the relationship between the visual and auditory spatial scenes. Signals related to eye movements affect the brain's auditory pathways from the ear through auditory cortex and beyond, but how these signals might contribute to computing the locations of sounds with respect to the visual scene is poorly understood. Here, we evaluated the information contained in the signals observed at the earliest processing stage, eye movement-related eardrum oscillations (EMREOs). We report that human EMREOs carry information about both horizontal and vertical eye displacement as well as initial/final eye position. We conclude that all of the information necessary to contribute to a suitable coordinate transformation of auditory spatial cues into a common reference frame with visual information is present in this signal. We hypothesize that the underlying mechanism causing EMREOs could impose a transfer function on any incoming sound signal, which could permit subsequent processing stages to compute the positions of sounds in relation to the visual scene.
Content may be subject to copyright.
1
Evidence for a system in the auditory periphery that may contribute to
linking sounds and images in space
David LK Murphy1,3,4, Cynthia D King1-4, Stephanie N Schlebusch1-4, Christopher A Shera7,
Jennifer M Groh1-6
1Department of Psychology and Neuroscience
2Department of Neurobiology
3Center for Cognitive Neuroscience
4Duke Institute for Brain Sciences
5Department of Computer Science
6Department of Biomedical Engineering
Duke University
7Dept. of Otolaryngology
University of Southern California
Acknowledgments
1
2
We are grateful to Dr. Matthew Cooper, Dr. Kurtis Gruters, Dr. David Kaylie, Dr. Jeff Mohl,
3
Meredith Schmehl, Dr. David Smith, Chloe Weiser and Shawn Willett for discussions and
4
assistance concerning this project. This work was supported by NIH (NIDCD) grant
5
DC017532 to JMG.
6
7
was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprint (whichthis version posted July 19, 2020. . https://doi.org/10.1101/2020.07.19.210864doi: bioRxiv preprint
2
Abstract
8
Eye movements alter the relationship between the visual and auditory spatial scenes.
9
Signals related to eye movements affect the brain’s auditory pathways from the ear
10
through auditory cortex and beyond, but how these signals might contribute to
11
computing the locations of sounds with respect to the visual scene is poorly understood.
12
Here, we evaluated the information contained in the signals observed at the earliest
13
processing stage, eye movement-related eardrum oscillations (EMREOs). We report
14
that human EMREOs carry information about both horizontal and vertical eye
15
displacement as well as initial/final eye position. We conclude that all of the information
16
necessary to contribute to a suitable coordinate transformation of auditory spatial cues
17
into a common reference frame with visual information is present in this signal. We
18
hypothesize that the underlying mechanism causing EMREOs could impose a transfer
19
function on any incoming sound signal, which could permit subsequent processing
20
stages to compute the positions of sounds in relation to the visual scene.
21
Introduction
22
Every time we move our eyes, our retinae move in relation to our ears. These
23
movements shift the relationship of the visual scene (as detected by the retinal surface)
24
and the auditory scene (as detected based on timing, loudness, and frequency
25
differences in relation to the head and ears). Precise information about the magnitude,
26
direction, starting/ending position, and time of occurrence of each eye movement is
27
therefore needed to connect the brain’s views of the visual and auditory scenes to one
28
another. While there is ample evidence that the brain uses information about eye
29
movements to support visual-auditory integration [1-3], and eye movement signals are
30
known to be rife within the auditory pathway [4-11] as well as association cortex and
31
oculomotor areas [12-23], the computational mechanisms that underlie this process are
32
poorly understood. In particular, it is not well known how the initial peripheral stages of
33
sensory signal processing contribute to this process. In hearing (as in other senses),
34
the brain exerts descending control over sensory transduction [24-29]. Thus, the brain
35
has the opportunity to tailor how ascending sensory signals are first encoded, and
36
potentially adjust ascending auditory signals to facilitate linking to the visual system at
37
later stages.
38
was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprint (whichthis version posted July 19, 2020. . https://doi.org/10.1101/2020.07.19.210864doi: bioRxiv preprint
3
Here, we assessed the potential contributions of one such process. Eye
39
movement-related eardrum oscillations (EMREOs) are a recently discovered
40
phenomenon in which the eardrum oscillates, time-locked to both saccade onset and
41
offset, in the absence of incoming sound [30]. These oscillations likely reflect
42
descending eye movement-related neural commands that influence the ears’ internal
43
motor elements - the middle ear muscles, outer hair cells, or some combination in a
44
still unknown fashion. In this study, we evaluated the information contained in this
45
signal in human participants. If the mechanism(s) that cause EMREOs are to adjust
46
ascending auditory signals to facilitate localizing sounds in a reference frame shared
47
with vision, these oscillations should vary with both horizontal and vertical dimensions
48
and should depend not only on eye displacements but also at least to a degree on
49
absolute eye positions.
50
We find that EMREOs consist of a basic waveform time-locked to the onset and
51
offset of eye movements that is modified by the spatial parameters of the eye
52
movements. Displacement of the eyes in the horizontal (i.e. contraversive vs.
53
ipsiversive) dimension has the largest impact, but the vertical dimensions and absolute
54
initial position of the eyes also contribute. The spatial signals contained in EMREOs are
55
sufficiently well-elaborated that target location can be predicted or “read out” from the
56
microphone signals recorded in the ear canal. These findings suggest that the eye-
57
movement information needed to accomplish a coordinate transformation on incoming
58
sounds is fully available in the most peripheral part of the auditory system. While the
59
precise mechanism that creates EMREOs remains unknown, we propose that that
60
underlying mechanism might introduce a transfer function to the sound transduction
61
process, serving to adjust the gain, latency, and/or spectral dependence of responses in
62
the cochlea. In principle, this could allow later stages of processing to extract an eye-
63
centered signal of sound location for registration with the eye-centered visual scene [1].
64
65
was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprint (whichthis version posted July 19, 2020. . https://doi.org/10.1101/2020.07.19.210864doi: bioRxiv preprint
4
66
Figure 1. Events of the tasks in time and space. a. Task events across time. Each trial
67
began with the onset of an initial fixation cue (labeled fixation). Participants made
68
saccades to the fixation point, then maintained fixation for a minimum of 200 ms. The
69
fixation cue was turned off and a new target requiring a second saccade was turned on.
70
Participants then fixated on the second target (labeled target) for another 200 ms, at which
71
point the target turned red indicating that the trial was over. The ear-canal recordings
72
were analyzed in conjunction with the second saccade. b-d. Spatial layouts of fixations
73
(black) and targets (green) for the three task designs used in this study.
74
Results
75
We used earbud microphones to record internally-generated sounds in the ear
76
canals of 10 human subjects while they performed eye-movement tasks involving
77
various visual fixation and target configurations, but no external sounds presented
78
(Figure 1). The events of the task are shown in time in Figure 1a. At the beginning of
79
each trial, the subject made a saccade (rapid eye movement) to a visual fixation target
80
(labeled fixation). After a period of 750ms of fixation presentation, a second target
81
appeared and the initial fixation target was turned off. Subjects then made a saccade to
82
that second target and maintained fixation for another 200 ms (labeled target). If
83
subjects did not maintain pre- and post-saccade fixation for at least 200ms, the trial was
84
discarded. The two periods of fixation ensured that the eyes were stable for a period
85
before and after the fixation-to-target saccade so baseline ear-canal recording
86
was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprint (whichthis version posted July 19, 2020. . https://doi.org/10.1101/2020.07.19.210864doi: bioRxiv preprint
5
measures could be established without eye movement. We analyzed the ear-canal
87
recordings associated with this second saccade.
88
Three different task designs are shown in panels b-c, and were chosen to focus
89
on three questions: 1) what are the respective contributions of relative eye
90
displacement vs absolute initial/final position of the eyes to the EMREO signal (three-
91
origin task)? 2) what are the respective contributions of purely horizontal vs purely
92
vertical saccade displacements to EMREOs (criss-cross task)? And 3) how do the
93
horizontal and vertical eye displacement components of EMREOs interact for saccades
94
spanning full 2-dimensional space, i.e., oblique saccades (grid task)? The answers to
95
these questions provide a sense of the full spatial dependence of these eardrum
96
oscillations.
97
Using the three-origin horizontal slice task, we first replicated our original findings
98
by analyzing only the trials that used similar targets as previously published work [30].
99
These consisted of a horizontal target array at a single azimuth or horizontal “slice” with
100
a fixed vertical offset from a single center fixation position. Then, we extended our
101
original research by including additional fixation positions as well as additional targets in
102
the analysis. Results for the conditions that most closely matched our previous study
103
are shown for an example subject’s left ear in Figure 2a-f. Panels a and b show
104
average horizontal eye position as a function of time for each of the targets between -24
105
and 24 degrees (the original study used -18 to +18 degrees, as well as a 6 degree
106
vertical offset compared to 9 degrees here). The traces are color coded by the target
107
location, with the inset showing the average eye traces in two dimensions. In panel a,
108
the saccades are aligned on saccade onset, defined as the first peak in the third
109
derivative of eye position over time (eye jerk), or the point in time when the change in
110
eye acceleration is highest. In panel b, the saccades are aligned on saccade offset,
111
defined as the second peak in eye jerk, or when the change in eye deceleration is
112
greatest. Panels c and d show the average velocity traces associated with these
113
saccades. Finally, panels e and f are the average microphone signals recorded in the
114
left ear at the time these saccades were made, color-coded by visual target.
115
was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprint (whichthis version posted July 19, 2020. . https://doi.org/10.1101/2020.07.19.210864doi: bioRxiv preprint
6
116
Figure 2. Results for one participant’s left ear for the three origin horizontal slice task.
117
The left column shows results aligned on saccade onset. The right column shows the
118
results aligned on saccade offset. Color corresponds to various target and fixation
119
combinations, as illustrated by the inset eye traces. Panels a-f involve only the center
120
initial fixation position and its secondary targets (replicating previous research). The
121
different shades of blue/green are ipsiversive eye movements and the red/yellow are
122
contraversive, regardless which ear is depicted. Panels g-h show results of microphone
123
recordings from all three initial fixation positions (12, 0, -12 degrees) and their
124
corresponding straight vertical saccades. All microphone data was z-scored relative to
125
the mean and standard deviation of a time window -150 to -120 ms prior to saccade
126
onset, i.e. during a period of steady fixation. Units on the y axis are therefore in units of
127
standard deviation relative to this baseline.
128
129
As described in our previous work, an internally-generated oscillation (i.e. sound)
130
occurs in conjunction with each eye movement. The phase and amplitude vary with the
131
was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprint (whichthis version posted July 19, 2020. . https://doi.org/10.1101/2020.07.19.210864doi: bioRxiv preprint
7
direction and magnitude of the saccade, and there is evident of phase-resetting at both
132
saccade onset and offset. The phase resetting at saccade offset is particularly apparent
133
in panel f, where the first peaks or troughs roughly line up with each other after offset,
134
whereas beforehand they are staggered according to the associated target.
135
The current study builds on our original findings by considering the contributions
136
of absolute initial/final eye position to the EMREO pattern. We evaluated the results for
137
all three fixation locations, the straight ahead fixation as well as 12 degrees on either
138
side, all in the 0 degree vertical plane. To assess the impact of initial/final eye position
139
on the EMREO signal, we compared ear-canal recordings from saccades that had the
140
same direction and amplitude, but different starting and ending positions. Panels 2g and
141
2h show an example of average microphone traces for three such fixation/target
142
combinations. Overall, the shape of the microphone trace appears similar but not
143
identical for the same eye displacement from different absolute positions.
144
To statistically quantify the pattern of dependence on eye position and
145
displacement, we deployed a regression analysis in which the microphone signal at
146
each moment in time Mic(t) was fit to the horizontal initial eye position (H) and change in
147
eye position (H) on that trial as follows:
148
Mic(t) = BH(t) H + BdH(t)H + C(t)
149
with BH and BdH expressing the contributions related to initial horizontal eye position and
150
eye displacement respectively, and C(t) providing a time-varying constant term that can
151
be thought of here as the average oscillation regardless of horizontal saccade metrics
152
(see Methods, equation 1). To eliminate a correlation between H and H, we included
153
only trials for which the horizontal target displacements occurred for all three initial
154
fixation positions, i.e. -12o to +12o targets for the 0o fixation position, -24o to 0o targets
155
for the -12o fixation, and 0o to 24o targets for the +12o fixation position.
156
Figure 3 shows the regression terms across time (aligned to saccade onset) for
157
the 10 subjects in our study. Red lines show the data for the right ear and blue for the
158
simultaneously recorded left ear. Thickened portions indicate periods of time when the
159
95% confidence intervals for the given regression coefficient did not include zero.
160
was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprint (whichthis version posted July 19, 2020. . https://doi.org/10.1101/2020.07.19.210864doi: bioRxiv preprint
8
The left column shows the regression coefficients associated with horizontal eye
161
displacement (BH(t)). These traces show a characteristic pattern that is present in all
162
subjects: a peak-trough-peak oscillation or trough-peak-trough oscillation depending
163
which ear is being recorded. The phase is inverted between ears, indicating that the
164
relevant aspect of eye displacement is whether it is ipsiversive or contraversive to the
165
ear canal where the signal is recorded. Every individual participant shows this
166
characteristic pattern. Other than modest differences in overall amplitude (note the
167
different calibration bars for each subplot), subject variability occurs mainly in the
168
duration and latency of the overall pattern. For example, the overall duration of the
169
oscillation is the longest in S13, lasting ~100 ms. Most subjects show a clear
170
peak/trough just prior to saccade onset (e.g. S05 and S71), while some do not (e.g.
171
S24).
172
The right column shows the constant term of the regression equation (C(t)),
173
reflecting the average oscillation that occurs regardless of the metrics of the associated
174
eye movement in the horizontal dimension. Here, the left and right ear traces are
175
extremely similar to each other in most individual subjects (except S17, who has a
176
magnitude difference across the two ears). Subjects differ from one another in the
177
timing/duration of this element of the effect. For example, S52 exhibits a fairly late
178
oscillation in the value of the constant term, beginning about 25 ms after saccade onset.
179
However, there is considerable consistency across subjects, with most showing a
180
trough-peak-trough pattern starting at saccade onset. (Note: All eye movements in this
181
task involved a vertical component, but that vertical component did not vary, so any
182
dependence of the EMREO on the vertical dimension would be captured by the
183
constant term in this version of our regression analysis. We will return to this issue
184
below using the criss-cross and grid tasks to provide a resolution.)
185
The middle column shows the corresponding data for the horizontal initial/final
186
eye position dependence. The microphone signal’s dependence on starting/ending eye
187
position was smaller and more variable across subjects than either of the other two
188
regression components. When present, it exhibited a counter-phase pattern across the
189
two ears similar to that seen for the horizontal displacement dependence.
190
191
was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprint (whichthis version posted July 19, 2020. . https://doi.org/10.1101/2020.07.19.210864doi: bioRxiv preprint
9
192
Figure 3. Regression results for all participants for the three origin horizontal slice task.
193
Thick line segments indicate epochs of time in which the corresponding regression term
194
was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprint (whichthis version posted July 19, 2020. . https://doi.org/10.1101/2020.07.19.210864doi: bioRxiv preprint
10
differed from zero (p<0.05). Percent values reflect the percentage of time points in the
195
window -5 ms to 70 ms with respect to saccade onset (the significance assessment
196
window) that differed from zero with 95% confidence. Y axis units are standard
197
deviations per degree (horizontal eye displacement and horizontal initial eye position
198
columns) or simply standard deviations (“constant” column) and vary across panels as
199
indicated by the calibration bars.
200
Quantifying the statistical significance at the individual subject level is non-trivial
201
because the microphone signal at a given time point is not independent of the signal at
202
adjacent time points. To provide a general sense of significance, we computed the
203
percentage of individual time points for which the relevant regression terms differed
204
from zero at the 95% confidence intervals during a time period of 5 ms before to 70 ms
205
after saccade onset, dubbed the significance assessment window, as indicated by the
206
horizontal bars in the top row of graphs. This range was chosen to roughly capture the
207
full oscillations of all subjects. Note that this percentage cannot reach 100% because
208
microphone signals oscillate around zero and thus the 95% confidence intervals will
209
include zero during those times. However, it provides a useful benchmark. The
210
percent-significant-time points for each individual subject’s left and right ear traces are
211
shown on each individual panel. For the constant terms, these percentages range from
212
15-86%, and for the horizontal displacement terms, they range from 13-82%. Using a
213
decision criterion that each individual subject must show at least 10% significant points
214
for at least one ear, all 10 subjects easily surpassed this and can be said to exhibit a
215
significant dependence of the microphone signal on horizontal eye displacement as well
216
as showing a significant constant term.
217
The pattern is weaker for the horizontal initial/final position component: 9 of 10
218
reached the criterion (except S11), but it should be noted that two do so just barely
219
(S17, S45). We conclude that on average there does exist a dependence on horizontal
220
initial/final position but that this dependence is weaker, more variable, and potentially
221
less regular than the dependence on horizontal eye displacement. The irregular nature
222
of the impact of this factor suggests that it may require methods other than regression to
223
fully characterize.
224
We next explored eye displacement sensitivity in two dimensions. In the criss-
225
cross task (Figure 1c), we assessed purely horizontal and purely vertical saccades
226
initiated from a single central fixation point. Figure 4 shows the results from the same
227
was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprint (whichthis version posted July 19, 2020. . https://doi.org/10.1101/2020.07.19.210864doi: bioRxiv preprint
11
example subject shown in Figure 2. The top four panels (4a-d) show the average
228
horizontal eye position and velocity traces for a slice of horizontal targets with no
229
vertical offset and with a finer target spacing of 3 degrees, aligned on saccade onset
230
(left panels) and offset (right panels). The corresponding microphone signals are shown
231
in panels 4e and f. The results are reasonably similar to our original findings, which
232
used a non-zero vertical offset (of the target from the initial eye position), confirming that
233
the previously observed horizontal displacement dependence applies even in the
234
absence of a vertical component to the saccade. The bottom 6 panels (4g-l) show the
235
results for purely vertical displacements to targets. Again, there is a clear oscillation
236
associated with vertical saccades, with this subject showing a larger signal for upward
237
than for downward saccades. The oscillation continues for approximately 50 ms or more
238
after the saccade has ended.
239
To fit this data statistically, we again deployed a regression model, this time
240
omitting the term for the horizontal initial/final position and instead incorporating a term
241
for the vertical displacement:
242
Mic(t) = BdH(t)
H + BdV(t)

V + C(t)
243
The trials with horizontal targets and the trials with vertical targets were included in the
244
same regression fit as they were collected in the same sessions. Figure 5 shows the
245
results for all subjects, in a similar format as Figure 3.
246
The dependence of the microphone signal on the horizontal displacement for
247
purely horizontal saccades (left column) is quite similar to those seen in our original
248
study [30] as well as to the three-origin horizontal slice task results presented in Figure
249
3. Likewise, the constant component (right column) is also similar overall. All 10
250
subjects meet criteria for showing a significant overall effect for each of these factors.
251
252
was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprint (whichthis version posted July 19, 2020. . https://doi.org/10.1101/2020.07.19.210864doi: bioRxiv preprint
12
253
Figure 4. Results from one participant’s left ear (same participant as in Figure 2) in the
254
criss-cross task. Conventions similar to Figure 2. Panels a-f are results for purely
255
horizontal saccades and panels g-l are results for purely vertical saccades.
256
was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprint (whichthis version posted July 19, 2020. . https://doi.org/10.1101/2020.07.19.210864doi: bioRxiv preprint
13
257
Figure 5. Regression results for all participants for the criss-cross task. Conventions
258
similar to Figure 3.
259
was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprint (whichthis version posted July 19, 2020. . https://doi.org/10.1101/2020.07.19.210864doi: bioRxiv preprint
14
The middle column shows the dependence on purely vertical saccade displacement.
260
Using our earlier established criterion (i.e. at least one ear exhibiting significance for this
261
factor 10% of the time), a vertical displacement signal is present in all 10 subjects.
262
However, the pattern of effects for vertical saccade amplitude was substantially more
263
variable across subjects and usually occurred much later than the effects for horizontal
264
saccade amplitude (see for example S52). Note that the sensitivity to vertical saccade
265
amplitude was in phase between ears in nearly all subjects, in contrast to the out-of-
266
phase sensitivity to horizontal saccade amplitude between ears.
267
Next we considered the question of how these distinct dependencies on
268
horizontal and vertical saccade amplitude interact with each other. How do the
269
EMREOs vary in full two-dimensional space? Can the microphone signal for a full
270
range of oblique saccades be predicted from the individual horizontal and vertical eye
271
movement displacement components? To evaluate this, we tested subjects on one
272
more task, a 2D grid of targets which incorporated combined horizontal and vertical
273
displacements from a single fixation origin (see Figure 1d). As shown in Figure 6, the
274
regression coefficient curves for horizontal displacement (BdH(t)) and the constant terms
275
(C(t)) were very similar in the two tasks (grand average across the population, black vs.
276
green traces). The vertical displacement term (BdV(t)) was more variable but still quite
277
similar across the two tasks.
278
279
was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprint (whichthis version posted July 19, 2020. . https://doi.org/10.1101/2020.07.19.210864doi: bioRxiv preprint
15
Figure 6. Grand average regression coefficient curves for the grid and criss-
280
cross tasks correspond well. Black and green lines represent the average coefficient
281
values across 10 subjects; shading reflects the standard error across subjects.
282
283
284
Figure 7. Combined results for all participants’ left ears for the grid task (black
285
lines), with predicted results from the criss-cross task (red lines). The traces indicate
286
the grand average of all the individual participants mean microphone signals, with the
287
shading indicating +/- the standard error across participants. Black values in
288
parentheses are the actual horizontal and vertical coordinates for each target in the grid
289
task. Corresponding red values indicate the inferred target location based on solving a
290
multivariate regression which fits the observed grid task microphone signals in the
291
significance assessment window (-5 to 70 ms with respect to saccade onset) to the
292
was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprint (whichthis version posted July 19, 2020. . https://doi.org/10.1101/2020.07.19.210864doi: bioRxiv preprint
16
observed regression weights from the criss-cross task for each target location. The
293
averages of these values in the horizontal and vertical dimensions are shown across the
294
top and right side.
295
Given the similarity of these results, we surmised that it should be possible to
296
predict the microphone signal observed in the grid task using the data from the criss-
297
cross task. Figure 7 shows the grand average microphone signal associated with each
298
targets in the grid task (black traces). The location of each trace corresponds to the
299
physical location of the associated target in the grid task. The superimposed predicted
300
wave forms (red traces) were generated from the BdH(t), BdV(t), and C(t) regression
301
coefficients fit to the criss-cross data, then evaluated at each target location and
302
moment in time to produce predicted curves for each of the locations tested in the grid
303
task.
304
Overall, there is good correspondence between the predicted and observed
305
curves, even for oblique saccade displacements that were not tested in the criss-cross
306
task, i.e. the top left, top right, bottom left, and bottom right corners of this figure. This
307
illustrates two things: a) the EMREO is reproducible across task contexts, and b) the
308
horizontal and vertical displacement dependencies interact in a reasonably linear way,
309
so that the signal observed for a combined horizontal-vertical saccade can be predicted
310
as the sum of the signals observed for straight-horizontal and straight-vertical saccades
311
with the corresponding component amplitudes.
312
Given that it is possible to predict the microphone signal from one task context to
313
another, it should be possible to infer the target and its associated eye movement from
314
the microphone signal itself. To do this, we again used the weights for the regression
315
equation
316
Mic(t) = BdH(t) H + BdV(t)V + C(t)
317
from the criss-cross data. We then used the Mic(t) values observed in the grid task to
318
solve this system of multivariate linear equations across the significance assessment
319
window (time interval -5 to 70 ms with respect to the saccade) to generate the “read out”
320
values of H andV associated with each target’s actual H and V. We conducted
321
this analysis on the left ear and right ear data separately. The left ear results of this
322
analysis are seen in each of the individual panels of Figure 7; the black values (e.g. -18,
323
was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprint (whichthis version posted July 19, 2020. . https://doi.org/10.1101/2020.07.19.210864doi: bioRxiv preprint
17
12) indicate the actual horizontal and vertical locations of the target, and the associated
324
red values indicate the inferred location of the target. Across the top of the figure, the
325
numbers indicate the average inferred horizontal location, and down the right side, the
326
numbers indicate the average inferred vertical location. These results indicate that, on
327
average, the targets can be read out in the proper order, but the spatial scale is
328
compressed: the average read-out values for the +/-18 degree horizontal targets are +/-
329
~11-12 degrees, and the averages for the vertical +/- 12 degree targets are +/- ~6-7
330
degrees.
331
We wondered whether the accuracy of this read-out could be improved by
332
combining signals recorded in each ear simultaneously given that it is known that the
333
brain uses binaural computations for reconstructing auditory space. We first considered
334
a binaural difference computation subtracting the right ear microphone recordings
335
from the left, eliminating the part of the signal that is common between the two ears.
336
Figure 8A illustrates the inferred location of the target (red dots) connected to the actual
337
location of the target (black dots). Generally, the horizontal dimension is well ordered
338
whereas the vertical dimension has considerable shuffling. This can also be seen in
339
panels B and C which show the relationship between the inferred target location and the
340
true target location broken out separately for each dimension. The correlation between
341
inferred and actual target is higher in the horizontal dimension than the vertical
342
dimension. This makes sense because the binaural difference computation serves to
343
diminish the contribution from aspects of the signal that are in phase across the two
344
ears, such as the dependence on vertical eye displacement. Improvement in the
345
vertical readout can be achieved by instead averaging the signals across the two ears
346
(D). A hybrid readout operation in which the horizontal location is computed from the
347
binaural difference, and the vertical location is computed from the binaural average,
348
produces a modest improvement in the overall reconstruction of target location.
349
Overall, these results parallel human sound localization which relies on a binaural
350
difference computation in the horizontal dimension (and is more accurate in that
351
dimension), vs. potentially monaural spectral cues for the vertical dimension (which is
352
less accurate) [31, 32].
353
354
was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprint (whichthis version posted July 19, 2020. . https://doi.org/10.1101/2020.07.19.210864doi: bioRxiv preprint
18
355
was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprint (whichthis version posted July 19, 2020. . https://doi.org/10.1101/2020.07.19.210864doi: bioRxiv preprint
19
Figure 8. Reading out target location from the grid task microphone signals based on
356
the regression coefficients from the criss-cross task. a. Inferred target location (red)
357
compared to actual target location (black), based on the difference between the two
358
ears’ microphone signals (left ear – right ear). b. Horizontal component of the read-out
359
target vs the actual horizontal component (left ear right ear microphone signals). c.
360
Same as (b) but for the vertical component. d. Vertical component computed from the
361
average of the two ears. e. A hybrid read-out model using binaural difference in the
362
horizontal dimension and binaural average in the vertical dimension.
363
Discussion
364
Sound locations are inferred from head-centered differences in sound arrival
365
time, loudness, and spectral content, but visual stimulus locations are inferred from eye-
366
centered retinal locations [31, 32]. Information about eye movements with respect to
367
the head/ears is critical for connecting the visual and auditory scenes to one another [1].
368
This insight has motivated a number of previous neurophysiological studies in various
369
brain areas in monkeys and cats, all of which showed that changes in eye position
370
affected the auditory response properties of at least some neurons (Inferior colliculus:
371
[7-11] ; auditory cortex: [4-6] ; superior colliculus: [17-23] ; frontal eye fields: [12, 33] ;
372
intraparietal cortex: [13-16]). These findings raised the question of where signals
373
related to eye movements first appear in the auditory processing stream. The discovery
374
of EMREOs [30] introduced the intriguing possibility that the computational process
375
leading to visual-auditory integration might begin in the most peripheral part of the
376
auditory system. Here we show that the signals present in the ear exhibit the properties
377
necessary for playing a role in this process: these signals are influenced by both the
378
horizontal and vertical components of eye movements, and display signatures related to
379
both change-in-eye-position and the absolute position of the eyes in the orbits.
380
Our present observations raise two key questions: what causes EMREOs and
381
how do those causes impact hearing/auditory processing? The proximate cause of
382
EMREOs is likely to be one or more of the known types of motor elements in the ear:
383
the middle ear muscles (stapedius and tensor tympani), which modulate the motion of
384
the ossicles [34-36], and the outer hair cells, which modulate the motion of the basilar
385
membrane [37]. One or more of these elements may be driven by descending brain
386
signals originating from within the oculomotor pathway and entering the auditory
387
pathway somewhere along the descending stream that ultimately reaches the ear via
388
was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprint (whichthis version posted July 19, 2020. . https://doi.org/10.1101/2020.07.19.210864doi: bioRxiv preprint
20
the 5th (tensor tympani), 7th (stapedius muscle), and/or 8th nerves (outer hair cells) [see
389
refs: 24, 25-29] for reviews). Efforts are currently underway in our laboratory to identify
390
EMREO generators/modulators [38, 39].
391
Uncovering the underlying mechanism should shed light on a second question:
392
does the temporal pattern of the observed EMREO signal reflect the time course and
393
nature of that underlying mechanism’s impact on auditory processing? It is not clear
394
how an oscillatory signal like the one observed here might contribute to hearing.
395
However, it is also not clear that the underlying effect is, in fact, oscillatory.
396
Microphones can only detect signals with oscillatory energy in the range of sensitivity of
397
the microphone. It is possible that the observed oscillations therefore reflect ringing
398
associated with a change in some mechanical property of the transduction system, and
399
that that change has a non-oscillatory temporal profile (Figure 9A). Of particular interest
400
would be a ramp-to-step profile in which aspects of the middle or inner ear shift from
401
one state to another during the course of a saccade and hold steady at the new state
402
during the subsequent fixation period. This kind of temporal profile would match the
403
time course of the saccade itself.
404
Available eye movement control signals in the oculomotor system include those
405
that follow this ramp-and-hold temporal profile, or tonic activity that is proportional to eye
406
position throughout periods of both movement and fixation. In addition to such tonic
407
signals, oculomotor areas also contain neurons that exhibit burst patterns, or elevated
408
discharge in association with the saccade itself, as well as combinations of burst and
409
tonic patterns [for reviews, see 40, 41]. It remains to be seen which of these signals or
410
signal combinations might be sent to the auditory periphery and where they might come
411
from. The paramedian pontine reticular formation (PPRF) is a strong candidate, having
412
been implicated in providing corollary discharge signals of eye movements in visual
413
experiments [42] [see also 43], and containing each of these basic temporal signal
414
profiles [40, 41]. Regardless of the source and nature of the descending corollary
415
discharge signal, the oscillations observed here should be thought of as possibly
416
constituting a biomarker for an underlying, currently unknown, mechanism, rather than
417
necessarily the effect itself.
418
was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprint (whichthis version posted July 19, 2020. . https://doi.org/10.1101/2020.07.19.210864doi: bioRxiv preprint
21
419
Figure 9. Temporal profiles of relevant signals and working conceptual model for how
420
EMREOs might relate to our ability to link visual and auditory stimuli in space. A.
421
Temporal profiles of signals. The EMREO is oscillatory whereas the eye movement to
422
which it is synchronized involves a ramp-and-hold temporal profile. Candidate source
423
neural signals in the brain might exhibit a ramp-and-hold (tonic) pattern, suggesting a
424
ramp-and-hold-like underlying effect on an as-yet-unknown peripheral mechanism, or
425
could derive from other known temporal profiles including bursts of activity time-locked
426
to saccades. B. Working conceptual model. The brain causes the eyes to move by
427
sending a command to the eye muscles. Each eye movement shifts the location of
428
was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprint (whichthis version posted July 19, 2020. . https://doi.org/10.1101/2020.07.19.210864doi: bioRxiv preprint
22
visual stimuli on the retinal surface. A copy, possibly a highly transformed one, of this
429
eye movement command is sent to the ear, altering ear mechanics in some unknown
430
way. When a sound occurs, the ascending signal to the brain will depend on the
431
combination of its location in head-centered space (based on the physical values of
432
binaural timing and loudness differences and spectral cues) and aspects of recent eye
433
movements and fixation position. This hybrid signal could then be read-out by the brain.
434
Despite these critical unknowns, it is useful to articulate a working conceptual
435
model of how EMREOs might facilitate visual and auditory integration (Figure 9b). The
436
general notion is that, by sending a copy of each eye movement command to the motor
437
elements of the auditory periphery, the brain keeps the ear informed about the current
438
orientation of the eyes. If, as noted above, these descending oculomotor signals cause
439
a ramp-to-step change in the state of tension time-locked to the eye movement and
440
lasting for the duration of each fixation period, they would effectively change the
441
transduction mechanism in an eye position/movement dependent fashion. In turn,
442
these changes could affect the latency, gain, or frequency-filtering properties of the
443
response to sound. The signal sent to the brain in response to an incoming sound would
444
ultimately reflect a mixture of the physical cues related to the location of the sound itself
445
- the ITD/ILD/spectral cues - and eye position/movement information.
446
Most neurophysiological studies report signals consistent with a hybrid code in
447
which information about sound location is blended in a complex fashion with information
448
about eye position and movement, both within and across neurons [5, 9, 10, 17, 18, 20,
449
33]. Computational modeling confirms that, in principle, these complex signals can be
450
“read out” to produce a signal of sound location with respect to the eyes [9]. However,
451
substantive differences do remain between the observations here and such neural
452
studies, chiefly in that the neural investigations have focused primarily on periods of
453
steady fixation. A more complete characterization of neural signals time-locked to
454
saccades is therefore needed [7, 44].
455
Note that this working model differs from a spatial attention mechanism in which
456
the brain might direct the ears to listen selectively to a particular location in space.
457
Rather, the response to sounds from any location, would be impacted by this eye
458
movement/position dependence in a consistent fashion across all sound locations.
459
Such a system may well work in concert with top-down attention, which has previously
460
was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprint (whichthis version posted July 19, 2020. . https://doi.org/10.1101/2020.07.19.210864doi: bioRxiv preprint
23
been shown to impact outer hair cells even when participants are required to fixate and
461
not make eye movements [45-51].
462
One question that has been explored previously concerns whether sound
463
localization is accurate or inaccurate when a brief sound is presented at the time of an
464
eye movement. Boucher et al [2] reported that perisaccadic sound localization is quite
465
accurate, which suggests that EMREOs (or their underlying mechanism) do not impair
466
perception. This is an important insight because given the frequency and duration of
467
eye movements - about 3/sec and with each associated EMREO signal lasting 100
468
ms or longer, it would be highly problematic if sounds could not be accurately detected
469
or localized when they occur in these time windows.
470
Finally, the findings from our present study argue against another theory: that
471
EMREOs might actually constitute the sounds of the eyeballs moving in their sockets,
472
transmitted to the ear canal via tissue/bone conduction and picked up with the
473
microphone. Several aspects of these results are inconsistent with this hypothesis.
474
The most obvious point is that the EMREO signal persists for many 10’s of ms after the
475
saccade has stopped. A second point is that the EMREO signal in a given ear depends
476
critically on the direction of the associated eye movement, but presumably any summed
477
sound of the left and right eyes moving in conjugate fashion is symmetric: a 10 degree
478
saccade involving both eye balls moving to the left and a 10 degree saccade with both
479
moving to the right should make the same aggregated eye-in-orbits sounds, but they
480
produce EMREO signals that are opposite in phase in each ear canal.
481
Overall, how brain-controlled mechanisms adjust the signaling properties of
482
peripheral sensory structures is critical for understanding sensory processing as a
483
whole. Auditory signals are known to adjust the sensitivity of the visual system via
484
sound-triggered pupil dilation [52], indicating that communication between these two
485
senses is likely to be a two-way street. The functional impact of such communication at
486
low-level stages is yet to be fully explored and may have implications for how individuals
487
compensate when the information from one sensory system is inadequate, either due to
488
natural situations such as noisy sound environments or occluded visual ones, or due to
489
physiological impairments in one or more sensory systems.
490
491
was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprint (whichthis version posted July 19, 2020. . https://doi.org/10.1101/2020.07.19.210864doi: bioRxiv preprint
24
Methods
492
General
493
Healthy human subjects that were 18 years of age or older with no known
494
hearing deficits or visual impairments beyond corrected vision were recruited from the
495
surrounding campus community (N=10; 5 female, 5 male). If subjects were unable to
496
perform the saccade task without vision correction, they were excluded from the study.
497
All study procedures involving subjects were approved by the Duke Institutional Review
498
Board, and all subjects received monetary compensation for their participation in the
499
study.
500
Acoustic signals in both ear canals were measured simultaneously with Etymotic
501
ER10B+ microphone systems coupled with ER2 earphones for sound stimulus
502
presentation (Etymotic Research, Elk Grove Village, IL). A low-latency audio interface
503
(Focusrite Scarlett 2i2, Focusrite Audio Engineering Ltd., High Wycombe, UK) was used
504
for audio capture and playback through the Etymotic hardware at a sampling rate of
505
48kHz. Eye tracking was performed with an Eyelink 1000 system sampling at 1000Hz.
506
Stimulus presentation and data acquisition were controlled through custom scripts and
507
elements of The Psychophysics Toolbox in MATLAB, with visual stimuli presented on a
508
large LED monitor.
509
In all experiments, eye position and microphone data were recorded while
510
subjects performed silent, visually-guided saccade tasks. Experimental sessions were
511
carried out in a darkened, acoustically isolated chamber (IAC) made anechoic with the
512
addition of acoustic damping wall tiles (Sonex). Subjects were seated and used an
513
adjustable chin rest, positioned 70 cm from the screen, to minimize head movement.
514
Experimental sessions were subdivided into multiple runs, approximately 5 minutes
515
each. This provided subjects with the opportunity to take a brief break from the
516
experiment if needed to maintain alertness or to address discomfort from maintaining
517
their posture. Each run typically consisted of approximately 120 trials and saccade
518
targets were presented in pseudorandom order.
519
Before each experimental session, the eye-tracking system was calibrated using
520
the calibration routine provided with the Eyelink system to register raw eye-tracking data
521
to gaze locations on the stimulus presentation screen. If the subject requested an
522
adjustment to the chin rest or left the recording chamber for a break, the eye-tracking
523
was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprint (whichthis version posted July 19, 2020. . https://doi.org/10.1101/2020.07.19.210864doi: bioRxiv preprint
25
calibration was repeated. Before each run, the microphone system was calibrated to
524
ensure that each microphone had a frequency response that was similar to the pre-
525
recorded frequency response of the microphones when placed in a volume that
526
approximated the geometry of the human ear canal (see supplementary methods). It
527
consisted of a 3ml syringe cut to accept the Etymotic earpieces. The syringe stopper
528
was pulled to 1.25 cm3 to approximate the volume of the human ear canal. A small
529
amount of gauze (.25cm3) was added to the volume to emulate the attenuation caused
530
by the soft tissue of the ear canal. The calibration routine played tones from 10 to
531
1600Hz, at a constant system output amplitude. As the purpose of this calibration was
532
to compare microphone function in a control volume with that in an earpiece just placed
533
in a subject, the weighting of the tones was not specifically calibrated. If the input-output
534
results of the same tone sequences were consistent between ears and matched the
535
overall shape of the syringe calibration curves, microphone placement was considered
536
successful. No sounds were delivered during periods of experimental data collection.
537
538
Task descriptions
539
All visually-guided saccade tasks followed the same stimulus timing sequence: initial
540
fixation points (referred to as fixation) were displayed on screen for 750ms and then
541
removed as the saccade targets (referred to as target) were presented for 750ms
542
(Figure 1a). Fixation and target locations were indicated by green dots that were 1.5cm
543
in diameter. Subjects were instructed to fixate on the initial fixation locations until targets
544
were presented on the screen, then to saccade to the targets and fixate on the targets
545
until they changed from green to red for the last 100ms of the target presentation at
546
which point the targets were removed from screen (the color change was intended to
547
help subjects maintain consistent fixation periods). Inter-trial-intervals were jittered
548
350±150ms. This was done to minimize the potential impact of non-saccade related
549
noise signals that may be periodic (i.e. heartbeat, external acoustic and electromagnetic
550
sources).
551
In the three origin horizontal slice task (Figure 1B), participants performed
552
saccades to multiple targets from three different initial eye positions at -12º, 0º, and
553
+12º horizontally and at 0º of elevation from central fixation. Nine saccade targets were
554
was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprint (whichthis version posted July 19, 2020. . https://doi.org/10.1101/2020.07.19.210864doi: bioRxiv preprint
26
located at 9º elevation and their horizontal locations ranged from -24º to +24º in 6º
555
degree increments
556
In the criss-cross task (Figure 1C), participants performed saccades to targets
557
along the vertical and horizontal axes from central fixation. Vertical targets ranged from
558
-15º to +15º in 3º increments and horizontal targets ranged from -18º to +18º in 3º
559
increments.
560
In the (Figure 1D), participants made saccades to 24 distinct targets of varying
561
vertical and horizontal placement combinations from a central fixation. Horizontal
562
location components ranged from -18º to +18º in 9º increments and vertical location
563
components ranged from -12º to +12º in 6º increments.
564
565
Preprocessing analysis
566
Saccade-microphone synchronization:
567
Microphone data was synchronized to two time points in the eye position data: the
568
onset and offset of the targeting saccade. These points in time were defined based on
569
the third derivative of eye position, or jerk. The first peak in the jerk represents the
570
moment when the change in the eye acceleration is greatest, and the second peak
571
represents the moment when the change in the eye deceleration is greatest. Prior to
572
each differentiation, a lowpass discrete filter with a 7ms window was used to smooth the
573
data and reduce the effects of noise and quantization error. This filter was normalized,
574
such that its output to a constant series of values equaled those values.
575
576
Trial exclusion criteria:
577
Trials were excluded based on saccade performance and the quality of recordings for
578
both microphone and eye tracking: 1) if subjects made a sequence of two or more
579
saccades to achieve the target; 2) if the eye tracking signal dropped out during the trial
580
(e.g. due to blinks); 3) if the eye movement was slow and drifting, rather than a
581
saccade; 4) if the saccade curved by more than 4.5º (subtended angle); or 5) subjects
582
failed to maintain 200ms of fixation before and after the saccade. If eye tracking
583
dropped samples that prevented the calculation of the saccade onset and offset times,
584
the trial was also excluded. On average these saccade-related exclusion criteria
585
resulted in the exclusion of about 12% of the trials.
586
was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprint (whichthis version posted July 19, 2020. . https://doi.org/10.1101/2020.07.19.210864doi: bioRxiv preprint
27
Prior to any further analysis, microphone data was downsampled from 48 kHz to
587
2 kHz sampling rate to reduce processing time given that the previously observed eye-
588
movement related signals of interest are well below 100 Hz [30]. Exclusion based on
589
noise in the microphone recordings was minimal. Within each block of trials, the mean
590
and standard deviation of the RMS values for each trial was calculated. Individual trials
591
were excluded if the microphone signal on that trial contained any individual values that
592
were more than 10 standard deviations away from that mean. This typically resulted in
593
the exclusion of < ~2% of the trials, after application of the eye position screen
594
described above.
595
596
Z scoring
597
To facilitate comparison across subjects, sessions, and experiments, all microphone
598
data reported in this study was z-scored within blocks and prior to the application of the
599
exclusion criteria described above. The mean and standard deviation of the
600
microphone values in a window -150 to -120 ms prior to saccade onset was used as the
601
normalization baseline period. When saccade onset- and offset-aligned data were
602
compared with each other (e.g. Figure 2), the same onset-aligned normalization window
603
was used for both.
604
605
Regression Analyses
606
Regression was used to assess how EMREOs vary with both eye position and
607
eye movement. The microphone signal at each moment in time Mic(t) was fit as
608
follows:
609
Mic(t) = BH(t) H + BdH(t)

H + BV(t) V + BdV(t)

V + C(t) (1)
610
where H and V correspond to the initial horizontal and vertical eye position and
H and
611
V correspond to the respective changes in position associated with that trial. The
612
slope coefficients BH, BdH, BV, BdV are time-varying and reflect the dependence of the
613
microphone signal on the respective eye movement parameters. The term C(t)
614
contributes a time-varying “constantindependent of eye movement metrics, and can be
615
was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprint (whichthis version posted July 19, 2020. . https://doi.org/10.1101/2020.07.19.210864doi: bioRxiv preprint
28
thought of as the best fitting average oscillation across all eye positions and
616
displacements.
617
The term C(t) was included for all regressions, but other parameters were
618
omitted when not relevant. For the grid task and criss-cross tasks, the model used
619
vertical and horizontal saccade displacement (BdH(t)

H , BdV(t)

V) as regression
620
variables but not BH(t) H or BV(t) V as initial position did not vary for those tasks. For the
621
three-origin task, the model included BH(t) H and BdH(t)

H but not the vertical
622
parameters because those did not vary for this task. However, note that this task did
623
involve a vertical change in position, so the C(t) term for this task would tend to
624
incorporate the contributions related to that fixed vertical component.
625
The regressions for all experiments were carried out twice: once with data
626
synchronized to saccade onset and once with data synchronized to saccade offset. The
627
analysis produced values for the intercept and variable weights (or slopes), their 95%
628
confidence intervals, R2, and p-value for each time point.
629
For the target readout analysis described in Figures 7 and 8, the horizontal and
630
vertical positions of the targets, rather than the associated eye movements, were used.
631
References
632
1. Groh, J.M. and D.L. Sparks, Two models for transforming auditory signals from head-centered to eye-
centered coordinates. Biological Cybernetics, 1992. 67(4): p. 291-302.
2. Boucher, L., J.M. Groh, and H.C. Hughes, Afferent delays and the mislocalization of perisaccadic
stimuli. Vision Research, 2001. 41(20): p. 2631-44.
3. Metzger, R.R., O.A. Mullette-Gillman, A.M. Underhill, Y.E. Cohen, and J.M. Groh, Auditory saccades
from different eye positions in the monkey: implications for coordinate transformations. Journal of
Neurophysiology, 2004. 92(4): p. 2622-7.
4. Fu, K.M., A.S. Shah, M.N. O'Connell, T. McGinnis, H. Eckholdt, P. Lakatos, J. Smiley, and C.E.
Schroeder, Timing and laminar profile of eye-position effects on auditory responses in primate auditory
cortex. J Neurophysiol, 2004. 92(6): p. 3522-31.
5. Maier, J.X. and J.M. Groh, Comparison of gain-like properties of eye position signals in inferior
colliculus versus auditory cortex of primates. Frontiers in Integrative Neuroscience, 2010. 4: p. 121-132.
6. Werner-Reiss, U., K.A. Kelly, A.S. Trause, A.M. Underhill, and J.M. Groh, Eye position affects activity in
primary auditory cortex of primates. Current Biology, 2003. 13: p. 554-562.
7. Bulkin, D.A. and J.M. Groh, Distribution of eye position information in the monkey inferior colliculus.
Journal of Neurophysiology, 2012. 107(3): p. 785-95.
8. Bulkin, D.A. and J.M. Groh, Distribution of visual and saccade related information in the monkey inferior
colliculus. Front Neural Circuits, 2012. 6: p. 61.
was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprint (whichthis version posted July 19, 2020. . https://doi.org/10.1101/2020.07.19.210864doi: bioRxiv preprint
29
9. Groh, J.M., A.S. Trause, A.M. Underhill, K.R. Clark, and S. Inati, Eye position influences auditory
responses in primate inferior colliculus. Neuron, 2001. 29(2): p. 509-518.
10. Porter, K.K., R.R. Metzger, and J.M. Groh, Representation of eye position in primate inferior colliculus.
Journal of Neurophysiology, 2006. 95(3): p. 1826-42.
11. Zwiers, M.P., H. Versnel, and A.J. Van Opstal, Involvement of monkey inferior colliculus in spatial
hearing. J Neurosci, 2004. 24(17): p. 4145-56.
12. Russo, G.S. and C.J. Bruce, Frontal eye field activity preceding aurally guided saccades. Journal of
Neurophysiology, 1994. 71(3): p. 1250-3.
13. Mullette-Gillman, O.A., Y.E. Cohen, and J.M. Groh, Motor-related signals in the intraparietal cortex
encode locations in a hybrid, rather than eye-centered, reference frame. Cerebral Cortex, 2009. 19(8):
p. 1761-75.
14. Mullette-Gillman, O.A., Y.E. Cohen, and J.M. Groh, Eye-centered, head-centered, and complex coding
of visual and auditory targets in the intraparietal sulcus. Journal of Neurophysiology, 2005. 94(4): p.
2331-52.
15. Cohen, Y.E. and R.A. Andersen, Reaches to sounds encoded in an eye-centered reference frame.
Neuron, 2000. 27(3): p. 647-52.
16. Stricanne, B., R.A. Andersen, and P. Mazzoni, Eye-centered, head-centered, and intermediate coding
of remembered sound locations in area LIP. Journal of Neurophysiology, 1996. 76(3): p. 2071-6.
17. Lee, J. and J.M. Groh, Auditory signals evolve from hybrid- to eye-centered coordinates in the primate
superior colliculus. Journal of Neurophysiology, 2012. 108(1): p. 227-42.
18. Jay, M.F. and D.L. Sparks, Auditory receptive fields in primate superior colliculus shift with changes in
eye position. Nature, 1984. 309(5966): p. 345-7.
19. Jay, M.F. and D.L. Sparks, Sensorimotor integration in the primate superior colliculus. I. Motor
convergence. J Neurophysiol, 1987. 57(1): p. 22-34.
20. Jay, M.F. and D.L. Sparks, Sensorimotor integration in the primate superior colliculus. II. Coordinates of
auditory signals. J Neurophysiol, 1987. 57(1): p. 35-55.
21. Zella, J.C., J.F. Brugge, and J.W. Schnupp, Passive eye displacement alters auditory spatial receptive
fields of cat superior colliculus neurons. Nat Neurosci, 2001. 4(12): p. 1167-9.
22. Populin, L.C., D.J. Tollin, and T.C. Yin, Effect of eye position on saccades and neuronal responses to
acoustic stimuli in the superior colliculus of the behaving cat. J Neurophysiol, 2004. 92(4): p. 2151-67.
23. Hartline, P.H., R.L. Vimal, A.J. King, D.D. Kurylo, and D.P. Northmore, Effects of eye position on
auditory localization and neural representation of space in superior colliculus of cats. Exp Brain Res,
1995. 104(3): p. 402-8.
24. Guinan, J.J., Jr., Cochlear mechanics, otacoustic emisssions, and medial olivocochlear efferents:
Twenty years of advances and controversies along with areas ripe for new work., in Perspectives on
Auditory Research. 2014, Springer: New York. p. 229-246.
25. Guinan, J.J., Jr., Olivocochlear efferents: anatomy, physiology, function, and the measurement of
efferent effects in humans. Ear Hear, 2006. 27(6): p. 589-607.
26. Mukerji, S., A.M. Windsor, and D.J. Lee, Auditory brainstem circuits that mediate the middle ear muscle
reflex. Trends Amplif, 2010. 14(3): p. 170-91.
27. Cooper, N.P. and J.J. Guinan, Jr., Efferent-mediated control of basilar membrane motion. J Physiol,
2006. 576(Pt 1): p. 49-54.
was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprint (whichthis version posted July 19, 2020. . https://doi.org/10.1101/2020.07.19.210864doi: bioRxiv preprint
30
28. Liberman, M.C. and J.J. Guinan, Jr., Feedback control of the auditory periphery: anti-masking effects of
middle ear muscles vs. olivocochlear efferents. J Commun Disord, 1998. 31(6): p. 471-82; quiz 483;
553.
29. Galambos, R., Suppression of auditory nerve activity by stimulation of efferent fibers to cochlea. J
Neurophysiol, 1956. 19(5): p. 424-37.
30. Gruters, K.G., D.L.K. Murphy, C.D. Jenson, D.W. Smith, C.A. Shera, and J.M. Groh, The eardrums
move when the eyes move: A multisensory effect on the mechanics of hearing. Proc Natl Acad Sci U S
A, 2018. 115 (6): p. E1309-E1318.
31. Blauert, J., Spatial hearing. 1997, Cambridge, MA: MIT Press.
32. Groh, J.M., Making space: how the brain knows where things are. 2014, Cambridge, MA: Harvard
University Press.
33. Caruso, V.C., D.S. Pages, M.A. Sommer, and J.M. Groh, Compensating for a shifting world: A
quantitative comparison of the reference frame of visual and auditory signals across three multimodal
brain areas. bioRxiv, 2019: p. 669333.
34. Hung, I.J. and P. Dallos, Study of the acoustic reflex in human beings. I. Dynamic characteristics. J
Acoust Soc Am, 1972. 52(4): p. 1168-80.
35. Mendelson, E.S., A sensitive method for registration of human intratympanic muscle reflexes. J Appl
Physiol, 1957. 11(3): p. 499-502.
36. Gelfand, S.A., The contralateral acoustic reflex. , in The acoustic reflex: Basic principles and clinical
applications 1984, Academic Press: New York, NY. p. 137-186.
37. Brownell, W.E., C.R. Bader, D. Bertrand, and Y. de Ribaupierre, Evoked mechanical responses of
isolated cochlear outer hair cells. Science, 1985. 227(4683): p. 194-6.
38. Schlebusch, S.N., M.W. Cooper, D.M. Kaylie, C.D. King, D.L.K. Murphy, C.A. Shera, and J.M. Groh,
Changes in saccade-related eardrum oscillations after surgical denervation of the stapedius muscle.
Soc Neurosci. Abstr., 2019.
39. Schlebusch, S.N., M.W. Cooper, D.M. Kaylie, C.D. King, D.L.K. Murphy, C.A. Shera, and J.M. Groh,
Changes in saccade-related eardrum oscillations after surgical denervation of the stapedius muscle. .
Association for Research in Otolaryngology Abstracts, 2020.
40. Takahashi, M. and Y. Shinoda, Brain Stem Neural Circuits of Horizontal and Vertical Saccade Systems
and their Frame of Reference. Neuroscience, 2018. 392: p. 281-328.
41. Fuchs, A.F., C.R. Kaneko, and C.A. Scudder, Brainstem control of saccadic eye movements. Annu Rev
Neurosci, 1985. 8: p. 307-37.
42. Sparks, D.L., L.E. Mays, and J.D. Porter, Eye movements induced by pontine stimulation: interaction
with visually triggered saccades. J Neurophysiol, 1987. 58(2): p. 300-18.
43. Guthrie, B.L., J.D. Porter, and D.L. Sparks, Corollary discharge provides accurate eye position
information to the oculomotor system. Science, 1983. 221(4616): p. 1193-5.
44. Porter, K.K., R.R. Metzger, and J.M. Groh, Visual- and saccade-related signals in the primate inferior
colliculus. Proceedings of the National Academy of Sciences of the United States of America, 2007.
104(45): p. 17855-60.
45. Delano, P.H., D. Elgueda, C.M. Hamame, and L. Robles, Selective attention to visual stimuli reduces
cochlear sensitivity in chinchillas. J Neurosci, 2007. 27(15): p. 4146-53.
46. Harkrider, A.W. and C.D. Bowers, Evidence for a cortically mediated release from inhibition in the
human cochlea. J Am Acad Audiol, 2009. 20(3): p. 208-15.
was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprint (whichthis version posted July 19, 2020. . https://doi.org/10.1101/2020.07.19.210864doi: bioRxiv preprint
31
47. Srinivasan, S., A. Keil, K. Stratis, K.L. Woodruff Carr, and D.W. Smith, Effects of cross-modal selective
attention on the sensory periphery: cochlear sensitivity is altered by selective attention. Neuroscience,
2012. 223: p. 325-32.
48. Srinivasan, S., A. Keil, K. Stratis, A.F. Osborne, C. Cerwonka, J. Wong, B.L. Rieger, V. Polcz, and D.W.
Smith, Interaural attention modulates outer hair cell function. Eur J Neurosci, 2014. 40(12): p. 3785-92.
49. Walsh, K.P., E.G. Pasanen, and D. McFadden, Selective attention reduces physiological noise in the
external ear canals of humans. I: auditory attention. Hear Res, 2014. 312: p. 143-59.
50. Walsh, K.P., E.G. Pasanen, and D. McFadden, Changes in otoacoustic emissions during selective
auditory and visual attention. J Acoust Soc Am, 2015. 137(5): p. 2737-57.
51. Wittekindt, A., J. Kaiser, and C. Abel, Attentional modulation of the inner ear: a combined otoacoustic
emission and EEG study. J Neurosci, 2014. 34(30): p. 9995-10002.
52. Bala, A.D. and T.T. Takahashi, Pupillary dilation response as an indicator of auditory discrimination in
the barn owl. J Comp Physiol [A], 2000. 186(5): p. 425-34.
was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprint (whichthis version posted July 19, 2020. . https://doi.org/10.1101/2020.07.19.210864doi: bioRxiv preprint
... The involvement of saccades in register with the attended loudspeaker location provides evidence for the existence of a reciprocal relationship as the recently discovered saccades induced eardrum oscillations 36,37 . That is, an auditory cue presentation at a particular location in space elicits oculomotor responses consistent with the sound origin. ...
... . While a sensory approach would argue for a direct effect of auditory processing on posterior regions (e.g.34 ), recent literature suggests that action-related sensory input mediate multisensory effects. For example, eye movements during auditory attention inform individual group differences within the dorsal attention network 35 and eye-movement related eardrum oscillations link sounds and images in space36,37 . Thus, alternatively, an affirmative case for the presence of saccades in register of auditory cue location might offer some explanation. ...
Preprint
Full-text available
It is well-established that power modulations of alpha oscillations (8-14 Hz) reflect the retinotopic organization of visuospatial attention. To what extend this organization generalizes to other sensory modalities is a topic of ongoing scientific debate. Here, we designed an auditory paradigm eliminating any visual input in which participants were required to attend to upcoming sounds from one of 24 loudspeakers arranged in a horizontal circular array around the head. Maintaining the location of an auditory cue was associated with a topographically modulated distribution of posterior alpha power resembling the findings known from visual attention. Alpha power modulations in all electrodes allowed us to predict the sound location in the horizontal plane using a forward encoding model. Importantly, this prediction was still possible, albeit weaker, when derived from the horizontal electrooculogram capturing saccadic behavior. We conclude that attending to an auditory target engages oculomotor and visual cortical areas in a topographic manner akin to the retinotopic organization associated with visual attention suggesting that the spatial distribution of alpha power reflects the supramodal organization of egocentric space.
... These oscillating movements of the eardrums, possibly the products of MEM contraction, were directionally consistent with directional eye movements. Indeed, further investigation has shown that eye movement-related eardrum oscillations (EMREOs) are a function of eye movement and absolute eye position ( Murphy et al., 2020 ). The mechanism of this interaction between sight and hearing has not been fully elucidated, however, should MEMs be found responsible for these EMREOs, the muscles would appear to have a spatial orientation function of some kind. ...
Article
The middle ear is a high-fidelity, broadband impedance transformer that transmits acoustic stimuli at the eardrum to the inner ear. It is home to the two smallest muscles in mammalian species, which modulate middle ear transmission. Of this pair, the function of the tensor tympani muscle (TTM) has remained obscure. We investigated the acoustic effects of this muscle in young adult gerbils. We measured changes in middle ear vibration produced by pulse-train-elicited TTM contraction – in the time-domain with a click stimulus and in the frequency-domain with multitone zwuis stimuli. In our click experiments, there was generally a small reduction in the primary peak of the response and a slight increase in the subsequent ringing, but there was essentially no change in the delay of the click response at the umbo (less than 1 μs change). In our multitone experiments, there were consistent patterns of attenuation and enhancement in the velocity responses at the umbo and ossicles. TTM contraction produced a narrow band of enhancement around 6 kHz (maximally ∼5dB) that can be modeled with an increased stiffness of an overdamped spring-mass resonance. At frequencies below 2 kHz and above 35 kHz, TTM contraction attenuated middle ear vibrations by as much as fivefold.
... 33,72 Further, inferential processing may play a role in the gradual tuning of perceptual systems for direct perception across development. Simple adaptations for perceiving cross-sensory magnitude or position information 73,74 or for adaptive signaling [75][76][77][78] may work in concert with learning, [79][80][81][82] language, 66 and cultural evolution 83 processes to support the development of taskrelevant representations. This arrangement could flexibly accommodate culture-specific emotion concepts and display rules. ...
Article
Emotionally expressive music and dance occur together across the world. This may be because features shared across the senses are represented the same way even in different sensory brain areas, putting music and movement in directly comparable terms. These shared representations may arise from a general need to identify environmentally relevant combinations of sensory features, particularly those that communicate emotion. To test the hypothesis that visual and auditory brain areas share a representational structure, we created music and animation stimuli with crossmodally matched features expressing a range of emotions. Participants confirmed that each emotion corresponded to a set of features shared across music and movement. A subset of participants viewed both music and animation during brain scanning, revealing that representations in auditory and visual brain areas were similar to one another. This shared representation captured not only simple stimulus features but also combinations of features associated with emotion judgments. The posterior superior temporal cortex represented both music and movement using this same structure, suggesting supramodal abstraction of sensory content. Further exploratory analysis revealed that early visual cortex used this shared representational structure even when stimuli were presented auditorily. We propose that crossmodally shared representations support mutually reinforcing dynamics across auditory and visual brain areas, facilitating crossmodal comparison. These shared representations may help explain why emotions are so readily perceived and why some dynamic emotional expressions can generalize across cultural contexts.
Article
Full-text available
Spatially selective modulation of alpha power (8–14 Hz) is a robust finding in electrophysiological studies of visual attention, and has been recently generalized to auditory spatial attention. This modulation pattern is interpreted as reflecting a top-down mechanism for suppressing distracting input from unattended directions of sound origin. The present study on auditory spatial attention extends this interpretation by demonstrating that alpha power modulation is closely linked to oculomotor action. We designed an auditory paradigm in which participants were required to attend to upcoming sounds from one of 24 loudspeakers arranged in a circular array around the head. Maintaining the location of an auditory cue was associated with a topographically modulated distribution of posterior alpha power resembling the findings known from visual attention. Multivariate analyses allowed the prediction of the sound location in the horizontal plane. Importantly, this prediction was also possible, when derived from signals capturing saccadic activity. A control experiment on auditory spatial attention confirmed that, in absence of any visual/auditory input, lateralization of alpha power is linked to the lateralized direction of gaze. Attending to an auditory target engages oculomotor and visual cortical areas in a topographic manner akin to the retinotopic organization associated with visual attention.
Article
Full-text available
Interactions between sensory pathways such as the visual and auditory systems are known to occur in the brain, but where they first occur is uncertain. Here, we show a multimodal interaction evident at the eardrum. Ear canal microphone measurements in humans (n = 19 ears in 16 subjects) and monkeys (n = 5 ears in three subjects) performing a saccadic eye movement task to visual targets indicated that the eardrum moves in conjunction with the eye movement. The eardrum motion was oscillatory and began as early as 10 ms before saccade onset in humans or with saccade onset in monkeys. These eardrum movements, which we dub eye movement-related eardrum oscillations (EMREOs), occurred in the absence of a sound stimulus. The amplitude and phase of the EMREOs depended on the direction and horizontal amplitude of the saccade. They lasted throughout the saccade and well into subsequent periods of steady fixation. We discuss the possibility that the mechanisms underlying EMREOs create eye movement-related binaural cues that may aid the brain in evaluating the relationship between visual and auditory stimulus locations as the eyes move.
Article
Full-text available
The inferior colliculus (IC) is an essential stop early in the ascending auditory pathway. Though normally thought of as a predominantly auditory structure, recent work has uncovered a variety of non-auditory influences on firing rate in the IC. Here, we map the location within the IC of neurons that respond to the onset of a fixation-guiding visual stimulus. Visual/visuomotor associated activity was found throughout the IC (overall, 84 of 199 sites tested or 42%), but with a far reduced prevalence and strength along recording penetrations passing through the tonotopically organized region of the IC, putatively the central nucleus (11 of 42 sites tested, or 26%). These results suggest that visual information has only a weak effect on early auditory processing in core regions, but more strongly targets the modulatory shell regions of the IC.
Article
Sensory signals for eye movements (visual and vestibular) are initially coded in different frames of reference but finally translated into common coordinates, and share the same final common pathway, namely the same population of extraocular motoneurons. From clinical studies in humans and lesion studies in animals, it is generally accepted that voluntary saccadic eye movements are organized in horizontal and vertical Cartesian coordinates. However, this issue is not settled yet, because neural circuits for vertical saccades remain unidentified. We recently determined brainstem neural circuits from the superior colliculus to ocular motoneurons for horizontal and vertical saccades with combined electrophysiological and neuroanatomical techniques. Comparing well-known vestibuloocular pathways with our findings of commissural excitation and inhibition between both superior colliculi, we proposed that the saccade system uses the same frame of reference as the vestibuloocular system, common semicircular canal coordinate. This proposal is mainly based on marked similarities (1) between output neural circuitry from one superior colliculus to extraocular motoneurons and that from a respective canal to its innervating extraocular motoneurons, (2) of patterns of commissural reciprocal inhibitions between upward saccade system on one side and downward system on the other, and between anterior canal system on one side and posterior canal system on the other, and (3) between the neural circuits of saccade and quick phase of vestibular nystagmus sharing brainstem burst neurons. In support of the proposal, commissural excitation of the superior colliculi may help to maintain Listing's law in saccades in spite of using semicircular canal coordinate.
Chapter
Reviewed are progress, exciting new developments, and areas that need work. Topics considered are: Cochlear amplification is from energy injected into the traveling wave by outer-hair-cell somatic motility. Calcium-activated stereocilia motility does not work at high frequencies because of the slowness of calcium binding and unbinding. Cochlear mechanics and micromechanics in the apical half of the cochlea are different from in the base. Cochlear micromechanics and the multiple fluid drives to inner hair cell (IHC) stereocilia. Interaction of the multiple IHC fluid drives can explain phase reversals in auditory nerve fiber responses without phase reversals in basilar membrane responses. The mechanisms by which medial olivocochlear (MOC) efferents change cochlear mechanics and micromechanics. The generation mechanisms for otoacoustic emissions (OAEs). Distortion product OAEs (DPOAEs) travel backward by slow traveling waves. Stimulus frequency OAEs (SFOAEs) arise mainly from near the peak of the traveling wave. Using OAEs to reveal cochlear properties. Cochlear tuning is sharper in humans than in cats, guinea pigs, and chinchillas. Measuring MOC effects using changes in OAEs and the need for high OAE signal-to-noise ratios. MOC effects in humans. The role of MOC efferents in hearing. MOC activity makes it easier to hear signals in noise. MOC activity and selective attention. MOC activity reduces acoustic trauma.
Article
Previous studies have demonstrated that the otoacoustic emissions (OAEs) measured during behavioral tasks can have different magnitudes when subjects are attending selectively or not attending. The implication is that the cognitive and perceptual demands of a task can affect the first neural stage of auditory processing—the sensory receptors themselves. However, the directions of the reported attentional effects have been inconsistent, the magnitudes of the observed differences typically have been small, and comparisons across studies have been made difficult by significant procedural differences. In this study, a nonlinear version of the stimulus-frequency OAE (SFOAE), called the nSFOAE, was used to measure cochlear responses from human subjects while they simultaneously performed behavioral tasks requiring selective auditory attention (dichotic or diotic listening), selective visual attention, or relative inattention. Within subjects, the differences in nSFOAE magnitude between inattention and attention conditions were about 2–3 dB for both auditory and visual modalities, and the effect sizes for the differences typically were large for both nSFOAE magnitude and phase. These results reveal that the cochlear efferent reflex is differentially active during selective attention and inattention, for both auditory and visual tasks, although they do not reveal how attention is improved when efferent activity is greater.
Article
Mounting evidence suggests that auditory attention tasks may modulate the sensitivity of the cochlea by way of the corticofugal and the medial olivocochlear (MOC) efferent pathways. Here, we studied the extent to which a separate efferent tract, the ‘uncrossed’ MOC, which functionally connects the two ears, mediates inter-aural selective attention. We compared distortion product otoacoustic emissions (DPOAEs) in one ear with binaurally presented primaries, using an intermodal target detection task in which participants were instructed to report the occurrence of brief target events (visual changes, tones). Three tasks were compared under identical physical stimulation: (i) report brief tones in the ear in which DPOAE responses were recorded; (ii) report brief tones presented to the contralateral, non-recorded ear; and (iii) report brief phase shifts of a visual grating at fixation. Effects of attention were observed as parallel shifts in overall DPOAE contour level, with DPOAEs relatively higher in overall level when subjects ignored the auditory stimuli and attended to the visual stimulus, compared with both of the auditory-attending conditions. Importantly, DPOAE levels were statistically lowest when attention was directed to the ipsilateral ear in which the DPOAE recordings were made. These data corroborate notions that top-down mechanisms, via the corticofugal and medial efferent pathways, mediate cochlear responses during intermodal attention. New findings show attending to one ear can significantly alter the physiological response of the contralateral, unattended ear, probably through the uncrossed-medial olivocochlear efferent fibers connecting the two ears.
Article
Attending to a single stimulus in a complex multisensory environment requires the ability to select relevant information while ignoring distracting input. The underlying mechanism and involved neuronal levels of this attentional gain control are still a matter of debate. Here, we investigated the influence of intermodal attention on different levels of auditory processing in humans. It is known that the activity of the cochlear amplifier can be modulated by efferent neurons of the medial olivocochlear complex. We used distortion product otoacoustic emission (DPOAE) measurements to monitor cochlear activity during an intermodal cueing paradigm. Simultaneously, central auditory processing was assessed by electroencephalography (EEG) with a steady-state paradigm targeting early cortical responses and analysis of alpha oscillations reflecting higher cognitive control of attentional modulation. We found effects of selective attention at all measured levels of the auditory processing: DPOAE levels differed significantly between periods of visual and auditory attention, showing a reduction during visual attention, but no change during auditory attention. Primary auditory cortex activity, as measured by the auditory steady-state response (ASSR), differed between conditions, with higher ASSRs during auditory than visual attention. Furthermore, the analysis of cortical oscillatory activity revealed increased alpha power over occipitoparietal and frontal regions during auditory compared with visual attention, putatively reflecting suppression of visual processing. In conclusion, this study showed both enhanced processing of attended acoustic stimuli in early sensory cortex and reduced processing of distracting input, both at higher cortical levels and at the most peripheral level of the hearing system, the cochlea.
Article
In this study, a nonlinear version of the stimulus-frequency OAE (SFOAE), called the nSFOAE, was used to measure cochlear responses from human subjects while they simultaneously performed behavioral tasks requiring, or not requiring, selective auditory attention. Appended to each stimulus presentation, and included in the calculation of each nSFOAE response, was a 30-ms silent period that was used to estimate the level of the inherent physiological noise in the ear canals of our subjects during each behavioral condition. Physiological-noise magnitudes were higher (noisier) for all subjects in the inattention task, and lower (quieter) in the selective auditory-attention tasks. These noise measures initially were made at the frequency of our nSFOAE probe tone (4.0 kHz), but the same attention effects also were observed across a wide range of frequencies. We attribute the observed differences in physiological-noise magnitudes between the inattention and attention conditions to different levels of efferent activation associated with the differing attentional demands of the behavioral tasks. One hypothesis is that when the attentional demand is relatively great, efferent activation is relatively high, and a decrease in the gain of the cochlear amplifier leads to lower-amplitude cochlear activity, and thus a smaller measure of noise from the ear.