Behavioral and objective measures of the precedence effect · Figure 2 Individual and mean results...
1
0 2 4 6 8 ï 0 2 4 6 8 10 12 14 ITD=300 us ICI (ms) /DJïZDYH 9 DPSOLWXGH UHGXFWLRQ G% 0 2 4 6 8 ITD=900 us ICI (ms) 0RQR /HIW 0RQR 5LJKW %LQDXUDO ABRs Aim • Compare ABRs to monaural and binaural stimulation to investigate whether binaural processes contribute to lag- suppression at brainstem level Stimuli Deviants with ICIs of 0-8 ms and ITDs of 300 and 900 μs were presented monaurally and binaurally. The ABRs were recorded using 4 surface electrodes: M1, M2, Fz (ground), Cz (reference). Left ear Right ear Results No evidence for a binaural contribution to lag-suppression at the brainstem level. 0 2 4 6 8 SV KE KG JC EC NL FB AL Mean Echoïthreshold (ms) Fusion Binaural MonoR MonoL Behavioral and objective measures of the precedence effect Federica Bianchi, Sarah Verhulst, Torsten Dau Centre for Applied Hearing Research, Technical University of Denmark Abstract The precedence effect (PE) refers to the dominance of directional information carried by a direct sound (lead) over the spatial information contained in its multiple reflections (lags) in sound localization. Many studies investigated the PE at different stages along the auditory pathway, but it is still unclear whether this perceptual phenomenon originates from peripheral or central processes. The present study aimed at investigating peripheral correlates of the PE, by comparing psychoacoustical data to objective measures of lag-suppression. The psychoacoustical data showed that the time range in which the PE operated (precedence window) was between 1 and 5 ms. Click-evoked otoacoustic emissions (CEOAEs) were recorded to investigate monaural lag-suppression at a peripheral stage of the auditory pathway, and showed largest lag-suppression within the precedence window. Auditory brainstem responses (ABRs) were used to investigate both monaural and binaural lag-suppression at the brainstem level. The ABRs to monaural stimulation reflected the peripheral lag-suppression, while the binaural ABRs did not show any additional contribution of binaural processes to the monaural suppression. The findings of this study demonstrate a monaural and peripheral component to lag-suppression, reflecting basilar membrane (BM) lead-lag impulse response interactions, in a time range from 1 to 4 ms. Introduction The precedence effect Consider a direct sound (lead) with a single reflection (lag). The lead-lag pair is reproduced in the free field by two clicks played over two loudspeakers at different locations. The perception of the lead-lag pair depends on the lead-lag delay and varies both in the number of perceived stimuli and in their location: Three perceptual phenomena occur within the precedence window: • fusion • lead-localization dominance PRECEDENCE • lag-discrimination suppression EFFECT Is the precedence effect a result of peripheral or central processes? Aim of the study To investigate the contribution of peripheral processing to the precedence effect by providing individual comparisons of objective and behavioral data. CEOAEs Aim • Investigate monaural lag-suppression at a peripheral stage of auditory processing Stimuli Deviants with ICIs of 0-8 ms and ITDs of 300 and 900 μs were presented using an interleaved technique. For each ICI and ITD condition, 1800 repetitions of the following three stimuli were presented: ✦ Single Click (SC) ✦ Double Click (DC) ✦ Double Click Inverted (DCI) Results Lead-lag interactions at BM level provide a peripheral source of monaural lag-suppression occurring for ICIs within the precedence window. Lead Lag Lead Lag Lead Lag SC Discussion Peripheral lag-suppression and echo-threshold ✦ For all six test subjects, the suppression for ICIs below the echo- threshold was significantly larger than the suppression above the echo- threshold. ✦ These results suggest that the lead-lag interactions on the BM level provide a peripheral source of lag-suppression occurring for ICIs within the precedence window. Monaural ABRs and CEOAE results The ABRs to monaural stimulation (blue and red dashed curves in Fig. 9) followed the trends of the CEOAE results (blue and red solid curves). However, the two curves were not expected to be identical: ✦ CEOAE results : reflect the amplitude-reduction of the backward traveling wave, which contains information of specific reflection sites along the basilar membrane [7, 8]. ✦ ABR results : reflect the neural activity elicited by the forward traveling wave, which should hypothetically reflect the synchronous activity of the whole basilar membrane. Even though both OAE and ABR results are expected to reflect monaural BM lead-lag interactions, the OAE only contains a subset of emission components present in the ABR. Thus, the shorter time range of suppression obtained in the ABR results might be due to the higher frequency content. Monaural and binaural ABRs ✦ For 4 out of 6 subjects, the lag-waveV reduction obtained for binaural stimulation was not larger than the reduction obtained for either monaural stimulations. ✦ These results suggest that binaural processes did not add any substantial contribution to monaural and peripheral suppression at the brainstem level, in agreement with [2]. Conclusion The results show that BM lead-lag interactions represent the main source of lag-suppression up to the brainstem. This suggests the existence of a monaural and peripheral component to the binaurally perceived PE. References [1] Litovsky RY (1999) The precedence effect. Journal of the Acoustical Society of America 106:1633-1654. [2] Damaschke J, Riedel H, Kollmeier B (2005) Neural correlates of the precedence effect in auditory evoked potentials. Hearing research 205:157-171. [3] Litovsky RY, Hawley ML, Colburn HS (1997) Measurement of precedence in monaural listeners. Meeting of the American Speech and Hearing Association, Boston, MA. [4] Kemp and Chum (1980) Properties of the generator of stimulated otoacoustic emissions. Hearing research 2: 213-232 [5] Kapadia S, Lutman ME (2000) Nonlinear temporal interactions in click-evoked otoacoustic emissions. Experimental data. Hearing research 146:89-100. [6] Verhulst S, Harte JM, Dau T (2011) Temporal suppression of the click-evoked otoacoustic emission level-curve. Journal of the Acoustical Society of America 129:1452- 1463. [7] Shera CA, Guinan JJ, Jr. (1999) Evoked otoacoustic emissions arise by two fundamentally different mechanisms: a taxonomy for mammalian OAEs. Journal of the Acoustical Society of America 105: 782-798. [8] Moleti et al.(2008) Transient evoked otoacoustic emission latency and estimates of cochlear tuning in preterm neonates. J. Acoust. Soc. Am. 124: 2984-2994. Observations: ✦ Mean echo-threshold occurs for an ICI of 4.3 ms (binaural stimulation). ✦ Similar thresholds for monaural and binaural stimulation. ✦ Fusion mechanism NOT dependent on binaural processes, in agreement with [3]. Figure 2 Individual and mean results of the fusion test. Echo- thresholds obtained for monaural and binaural stimulation. Figure 3 Results of the lateralization and ITD- detection tests for subject KE. The markers indicate the lateralization of the lead-lag pair as a function of ICI and ITD. The green highlighted area depicts the region below the lateralization threshold, where the percept of the lead-lag pair is centered. Observations: ✦ Higher ITD-thresholds for an ICI between 1-5 ms indicate stronger lead- dominance in this time window. ✦ Precedence window between 1 and 4.3 ms, where lead-dominance, lag- suppression and fusion occur. 0 1 2 3 4 5 6 7 8 0 200 400 600 800 1000 ICI (ms) ITD (us) Left Center Center & Center Center & Right Center & Left Center & Left & Right Lateralization thresh Detection thresh 1/3 response 2/3 responses 3/3 responses Figure 1 Perceived location of a lead-lag pair as a function of the lead-lag delay. The head schematics refers to a free field setup, where the leading speaker is towards the left and the lagging speaker towards the right. The time ranges are defined for click stimuli [1]. Derived suppressed response (DS) Unsuppressed response (US) Figure 5 Monaural and peripheral lag-suppression (dB) for subject KE (ITD=300 μs) as a function of the ICI. The dashed curve shows the monaural right lag-suppression, the solid curve the monaural left suppression. 10 0 1 2 3 4 5 6 7 8 9 R C L lead-lag delay (ms) Perceived Location Summing window Precedence Echo Echo- threshold window window Lead Lag Left ear Right ear Lead Lag ICI Reference Lead Lag Left ear Right ear Lead Lag ICI ITD Deviant Lead - Lag Figure 4 Procedure to derive the suppressed response from the recordings [4, 5], by removing the CEOAE component due to the leading click, while keeping the component due to the lagging click and the nonlinear component due to the ICI. 0 1 2 3 4 5 6 7 8 0 1 2 3 4 5 6 7 8 9 10 11 ICI (ms) /DJïVXSSUHVVLRQ G% 0RQR /HIW 0RQR 5LJKW Lead Lag ICI ITD L R echo-thresh Lag- suppression = Lrms (US) − Lrms (DS) Observations: ✦ The peripheral and monaural lag-suppression is largest for ICIs within the precedence window (in agreement with [5, 6]). ✦ Considering the linear coherent reflection theory of generation of CEOAEs [7], the peripheral lag-suppression reflects basilar membrane (BM) lead-lag impulse response interactions. 0 5 10 15 echoïthr dB Suppr (ITD = 300 us) MonoL MonoR Bin 0 2 4 6 8 0 5 10 15 echoïthr ICI (ms) dB Suppr (ITD = 900 us) 0 2 4 6 8 ICI (ms) 0 2 4 6 8 ICI (ms) OAE ABR OAE ABR ABR Bin ï0.3 0 0.3 ITD = 300 us ITD = 900 us ï0.3 0 0.3 ï0.3 0 0.3 ï0.3 0 0.3 voltage (uV) ï0.3 0 0.3 ï0.3 0 0.3 5 10 15 ï0.3 0 0.3 latency (ms) 5 10 15 latency (ms) ICI=0 ms ICI=1 ms ICI=2 ms ICI=3 ms ICI=4 ms ICI=5 ms ICI=8 ms Increasing ICI Figure 7 ABRs recorded for subject KE for binaural stimulation (left: ITD=300 μs; right: ITD=900 μs). The waves V elicited by lead and lag are indicated by downward-pointing triangles. Observations: ✦ Lag-wave V amplitude increases with increasing ICI (Fig. 7). ✦ Lag-wave V amplitude reductions for ICIs up to 4 ms. ✦ Binaural lag-wave V amplitude reduction does not show an additional reduction to the monaural one. Figure 8 Reduction of lag-wave V amplitude (dB) as a function of ICI (subject KE), for either monaural stimulations and for binaural stimulation (left: ITD=300 μs; right: ITD=900 μs) Figure 9 Lag-suppression [dB] as a function of ICI, obtained from CEOAE amplitude-reduction (blue and red solid curves) and ABR wave V-reduction (blue and red dashed curves, black curves), for monaural and binaural stimulation (subject KE). The monaural and binaural echo- thresholds, obtained in the fusion test, are represented by an arrow. Contact: Federica Bianchi, [email protected] Sarah Verhulst, [email protected] Psychoacoustical experiments Aim • Define the three windows of perception • Define the lateralization of the lead-lag pair for each window Stimuli Experiments Six normal hearing subjects performed the following 3 experiments: 1. Fusion test : define the echo-threshold for monaural and binaural stimulation (“1 click or 2 clicks?”). 2. ITD-detection test : define the minimum interaural time difference (ITD) for lateralization [2]. 3. ITD-lateralization test : specify the perception of the lead-lag pair as a function of the inter-click interval (ICI). Results Figure 6 Stimuli used in the ABR recordings and head schematics showing the position of the electrodes M1, M2 and Cz.
Transcript of Behavioral and objective measures of the precedence effect · Figure 2 Individual and mean results...
0 2 4 6 8
0
2
4
6
8
10
12
14
ITD=300 us
ICI (ms)0 2 4 6 8
ITD=900 us
ICI (ms)
ABRsAim• Compare ABRs to monaural and binaural stimulation to
investigate whether binaural processes contribute to lag-suppression at brainstem level
StimuliDeviants with ICIs of 0-8 ms and ITDs of 300 and 900 μs were presented monaurally and binaurally. The ABRs were recorded using 4 surface electrodes: M1, M2, Fz (ground), Cz (reference).
Left ear
Right ear
Results
No evidence for a binaural contribution to lag-suppression at the brainstem level.
0 2 4 6 8
SV
KE
KG
JC
EC
NL
FB
AL
Mean
Echo threshold (ms)
Fusion
BinauralMonoRMonoL
Behavioral and objective measures of the precedence effectFederica Bianchi, Sarah Verhulst, Torsten Dau
Centre for Applied Hearing Research, Technical University of Denmark
AbstractThe precedence effect (PE) refers to the dominance of directional information carried by a direct sound (lead) over the spatial information contained in its multiple reflections (lags) in sound localization. Many studies investigated the PE at different stages along the auditory pathway, but it is still unclear whether this perceptual phenomenon originates from peripheral or central processes. The present study aimed at investigating peripheral correlates of the PE, by comparing psychoacoustical data to objective measures of lag-suppression. The psychoacoustical data showed that the time range in which the PE operated (precedence window) was between 1 and 5 ms. Click-evoked otoacoustic emissions (CEOAEs) were recorded to investigate monaural lag-suppression at a peripheral stage of the auditory pathway, and showed largest lag-suppression within the precedence window. Auditory brainstem responses (ABRs) were used to investigate both monaural and binaural lag-suppression at the brainstem level. The ABRs to monaural stimulation reflected the peripheral lag-suppression, while the binaural ABRs did not show any additional contribution of binaural processes to the monaural suppression. The findings of this study demonstrate a monaural and peripheral component to lag-suppression, reflecting basilar membrane (BM) lead-lag impulse response interactions, in a time range from 1 to 4 ms.
IntroductionThe precedence effectConsider a direct sound (lead) with a single reflection (lag). The lead-lag pair is reproduced in the free field by two clicks played over two loudspeakers at different locations. The perception of the lead-lag pair depends on the lead-lag delay and varies both in the number of perceived stimuli and in their location:
Three perceptual phenomena occur within the precedence window:• fusion• lead-localization dominance PRECEDENCE • lag-discrimination suppression EFFECT
Is the precedence effect a result of peripheral or central processes?
Aim of the studyTo investigate the contribution of peripheral processing
to the precedence effect by providing individual comparisons of objective and behavioral data.
CEOAEsAim• Investigate monaural lag-suppression at a peripheral stage of
auditory processing
StimuliDeviants with ICIs of 0-8 ms and ITDs of 300 and 900 μs were presented using an interleaved technique. For each ICI and ITD condition, 1800 repetitions of the following three stimuli were presented:✦ Single Click (SC)✦ Double Click (DC)✦ Double Click Inverted (DCI)
Results
Lead-lag interactions at BM level provide a peripheral source of monaural lag-suppression occurring
for ICIs within the precedence window.
Lead Lag Lead
Lag
Lead Lag
SC
DiscussionPeripheral lag-suppression and echo-threshold✦ For all six test subjects, the suppression for ICIs below the echo-
threshold was significantly larger than the suppression above the echo-threshold.
✦ These results suggest that the lead-lag interactions on the BM level provide a peripheral source of lag-suppression occurring for ICIs within the precedence window.
Monaural ABRs and CEOAE resultsThe ABRs to monaural stimulation (blue and red dashed curves in Fig. 9) followed the trends of the CEOAE results (blue and red solid curves). However, the two curves were not expected to be identical:✦ CEOAE results: reflect the amplitude-reduction of the backward
traveling wave, which contains information of specific reflection sites along the basilar membrane [7, 8].
✦ ABR results: reflect the neural activity elicited by the forward traveling wave, which should hypothetically reflect the synchronous activity of the whole basilar membrane.
Even though both OAE and ABR results are expected to reflect monaural BM lead-lag interactions, the OAE only contains a subset of emission components present in the ABR. Thus, the shorter time range of suppression obtained in the ABR results might be due to the higher frequency content.
Monaural and binaural ABRs✦ For 4 out of 6 subjects, the lag-waveV reduction obtained for binaural
stimulation was not larger than the reduction obtained for either monaural stimulations.
✦ These results suggest that binaural processes did not add any substantial contribution to monaural and peripheral suppression at the brainstem level, in agreement with [2].
ConclusionThe results show that BM lead-lag interactions represent the main source of lag-suppression up to the brainstem. This suggests the existence of a monaural and peripheral component to the binaurally perceived PE.References[1] Litovsky RY (1999) The precedence effect. Journal of the Acoustical Society of America 106:1633-1654.[2] Damaschke J, Riedel H, Kollmeier B (2005) Neural correlates of the precedence effect in auditory evoked potentials. Hearing research 205:157-171.[3] Litovsky RY, Hawley ML, Colburn HS (1997) Measurement of precedence in monaural listeners. Meeting of the American Speech and Hearing Association, Boston, MA.[4] Kemp and Chum (1980) Properties of the generator of stimulated otoacoustic emissions. Hearing research 2: 213-232[5] Kapadia S, Lutman ME (2000) Nonlinear temporal interactions in click-evoked otoacoustic emissions. Experimental data. Hearing research 146:89-100.[6] Verhulst S, Harte JM, Dau T (2011) Temporal suppression of the click-evoked otoacoustic emission level-curve. Journal of the Acoustical Society of America 129:1452- 1463.[7] Shera CA, Guinan JJ, Jr. (1999) Evoked otoacoustic emissions arise by two fundamentally different mechanisms: a taxonomy for mammalian OAEs. Journal of the Acoustical Society of America 105: 782-798.[8] Moleti et al.(2008) Transient evoked otoacoustic emission latency and estimates of cochlear tuning in preterm neonates. J. Acoust. Soc. Am. 124: 2984-2994.
Observations:✦ Mean echo-threshold occurs for an ICI of 4.3 ms (binaural stimulation).✦ Similar thresholds for monaural and binaural stimulation.✦ Fusion mechanism NOT dependent on binaural processes, in
agreement with [3].
Figure 2 Individual and mean results of the fusion test. Echo-thresholds obtained for monaural and binaural stimulation.
Figure 3 Results of the lateralization and ITD-detection tests for subject KE. The markers indicate the lateralization of the lead-lag pair as a function of ICI and ITD. The green highlighted area depicts the region below the lateralization threshold, where the percept of the lead-lag pair is centered.
Observations:✦ Higher ITD-thresholds for an ICI between 1-5 ms indicate stronger lead-
dominance in this time window.✦ Precedence window between 1 and 4.3 ms, where lead-dominance, lag-
suppression and fusion occur.
0 1 2 3 4 5 6 7 8
0
200
400
600
800
1000
ICI (ms)
ITD
(us)
LeftCenterCenter & CenterCenter & RightCenter & LeftCenter & Left & RightLateralization threshDetection thresh1/3 response2/3 responses3/3 responses
Figure 1 Perceived location of a lead-lag pair as a function of the lead-lag delay. The head schematics refers to a free field setup, where the leading speaker is towards the left and the lagging speaker towards the right. The time ranges are defined for click stimuli [1].
Derived suppressed response (DS)
Unsuppressed response (US)
Figure 5 Monaural and peripheral lag-suppression (dB) for subject KE (ITD=300 μs) as a function of the ICI. The dashed curve shows the monaural right lag-suppression, the solid curve the monaural left suppression.
100 1 2 3 4 5 6 7 8 9
R
C
L
lead-lag delay (ms)
Perc
eive
d L
ocat
ion
Summingwindow
Precedence EchoEcho-thresholdwindow window
Lead Lag
Left ear
Right ear
Lead Lag
ICI
Reference
LeadLag
Left ear
Right ear
Lead Lag
ICI
ITD
Deviant
Lead - Lag
Figure 4 Procedure to derive the suppressed response from the recordings [4, 5], by removing the CEOAE component due to the leading click, while keeping the component due to the lagging click and the nonlinear component due to the ICI.
0 1 2 3 4 5 6 7 80123456789
1011
ICI (ms)
Lead Lag
ICI
ITDL
Recho-thresh
Lag- suppression = Lrms (US) − Lrms (DS)
Observations:✦ The peripheral and monaural lag-suppression is largest for ICIs within
the precedence window (in agreement with [5, 6]).✦ Considering the linear coherent reflection theory of generation of
CEOAEs [7], the peripheral lag-suppression reflects basilar membrane (BM) lead-lag impulse response interactions.
0
5
10
15
echo thr
ICI (ms)
dB S
uppr
(ITD
= 3
00 u
s)
MonoL
ICI (ms)
MonoR
ICI (ms)
Bin
0 2 4 6 8
0
5
10
15
echo thr
ICI (ms)
dB S
uppr
(ITD
= 9
00 u
s)
0 2 4 6 8ICI (ms)
0 2 4 6 8ICI (ms)
OAEABR
OAEABR
ABR Bin
0.30
0.3
ITD = 300 us ITD = 900 us
0.30
0.30.3
00.30.3
00.3
volta
ge (u
V)
0.30
0.30.3
00.3
5 10 15
0.30
0.3
latency (ms)5 10 15
latency (ms)
ICI=0 ms
ICI=1 ms
ICI=2 ms
ICI=3 ms
ICI=4 ms
ICI=5 ms
ICI=8 ms
Increasing ICIFigure 7 ABRs recorded for subject KE for binaural stimulation (left: ITD=300 μs; right: ITD=900 μs). The waves V elicited by lead and lag are indicated by downward-pointing triangles.
Observations:✦ Lag-wave V amplitude increases with increasing ICI (Fig. 7).✦ Lag-wave V amplitude reductions for ICIs up to 4 ms.✦ Binaural lag-wave V amplitude reduction does not show an additional
reduction to the monaural one.
Figure 8 Reduction of lag-wave V amplitude (dB) as a function of ICI (subject KE), for either monaural stimulations and for binaural stimulation (left: ITD=300 μs; right: ITD=900 μs)
Figure 9 Lag-suppression [dB] as a function of ICI, obtained from CEOAE amplitude-reduction (blue and red solid curves) and ABR wave V-reduction (blue and red dashed curves, black curves), for monaural and binaural stimulation (subject KE). The monaural and binaural echo-thresholds, obtained in the fusion test, are represented by an arrow.
Contact:Federica Bianchi, [email protected]
Sarah Verhulst, [email protected]
Psychoacoustical experimentsAim• Define the three windows of perception• Define the lateralization of the lead-lag pair for each window
Stimuli
ExperimentsSix normal hearing subjects performed the following 3 experiments:1. Fusion test: define the echo-threshold for monaural and binaural
stimulation (“1 click or 2 clicks?”).2. ITD-detection test: define the minimum interaural time difference (ITD)
for lateralization [2].3. ITD-lateralization test: specify the perception of the lead-lag pair as a
function of the inter-click interval (ICI).
Results
Figure 6 Stimuli used in the ABR recordings and head schematics showing the position of the electrodes M1, M2 and Cz.
![Secure Concurrent Constraint Programming - …catuscia/Talks/060820_ICLP/sccp.pdfRemark 1. (Behavioral Charecterization ... Notice thecorrespondencebetween (νx ... [Mil95] J. Millen.](https://static.fdocument.org/doc/165x107/5ab3d54a7f8b9ac3348ea67e/secure-concurrent-constraint-programming-catusciatalks060820iclpsccppdfremark.jpg)
