Exploring the Amplitudes of Binaural Interaction Components Elicited by Diverse Stimuli and Their Relationships With Behavioral Measures in Individuals With Normal Hearing
Article information
Abstract
Background and Objectives
This study was performed to comprehensively examine the amplitudes of the binaural interaction components (BICs) elicited by chirps, clicks, and 500 Hz tone-burst stimuli in individuals with normal hearing. Electrophysiological evidence of BICs was obtained and assessed for correlations with interaural time difference (ITD) and interaural level difference (ILD).
Subjects and Methods
Sixteen adults (4 males and 12 females) with normal hearing participated in this study. Auditory brainstem responses (ABRs) to chirp, click, and 500 Hz tone-burst stimuli were recorded, and BICs were derived based on wave V. The behavioral thresholds of ITDs and ILDs across multiple frequencies were obtained and analyzed.
Results
BICs were found in most participants, regardless of stimulus type. The amplitudes of BICs elicited by chirps were the highest, followed by those elicited by clicks and 500 Hz tone-bursts. A significant correlation was found between the amplitudes of chirp-evoked BICs and the thresholds of 500 Hz ITDs and ILDs.
Conclusions
This study found that chirp stimuli may be effective in eliciting BIC and predicting behavioral binaural interaction processing at low frequencies.
Introduction
The binaural interaction component (BIC) has been investigated as a method to assess binaural integration in the auditory brainstem [1,2]. Previously published studies on BIC have reported that binaural stimulation results in attenuated peak-to-peak amplitudes compared to the sum of two monaural conditions in auditory evoked potentials (AEP) measurements. These measurements include wave V in auditory brainstem response (ABR), Pa peak in middle latency response, and spectral amplitude changes in steady-state response [3]. This physiological phenomenon of attenuated responses is believed to arise from inhibitory activities across the central auditory nervous system (CANS), extending beyond the superior olivary complex (SOC) to the auditory cortex [4]. Specifically, researchers focused on the medial superior olive (MSO) and lateral superior olive (LSO) in the SOC as the first CANS structures to record binaural activities. Various types of neurons function to receive and transmit electrical signals through multiple afferent and efferent directions in these structures [3]. Tolnai and Klump [5] also reported evidence of binaural interaction in animal studies, isolating the target area in the SOC and concluding that inhibitory neurons in the MSO and LSO are the primary origins that attenuate contralateral signals. Some researchers proposed the exclusive function of MSO and LSO for interaural time difference (ITD) and interaural level difference (ILD) signals, while others have been unable to predict their contribution [6,7]. The amount of attenuation in the binaural condition is mathematically calculated as the sum of two monaural conditions minus the binaural condition: (L+R) – B [2]. In ABR, the peak-to-peak amplitude between positive wave V and successive negative peak is a prominent clue in understanding BIC changes.
A click stimulus has been commonly used to observe conventional ABR waveforms in clinical fields and research labs. As a transient onset and complex stimulus, a click can elicit strong short-latency neural responses because it stimulates a wide range of the basilar membrane in the cochlea simultaneously. However, the tonotopic organization in the cochlea results in longer traveling distances for lower-frequency components to reach their characteristic frequency at the apical end of the basilar membrane. As a result, imbalances between high and low frequencies of neural energy contribute to the ABR. Thus, even if a click stimulus contains a wide range of spectral cues, a click ABR practically yields high-frequency dominant responses, such as 2,000 Hz–4,000 Hz [8]. This limitation can introduce potential bias, especially for individuals with hearing loss at high frequencies. Consequently, binaural hearing evaluation might incline towards high-frequency ranges only [6].
On the other hand, the chirp stimulus, designed by Dau, et al. [8], was introduced to improve neural synchrony in auditory nerve fibers and enhance the balance of cochlear activation across the basilar [9]. The acoustical characteristic of the chirp stimulus is that low-frequency components are presented earlier than high-frequency components, authorizing low frequencies to reach their characteristic frequencies in the apical region of the basilar membrane [10]. This design results in a wider range of cochlear activities, simultaneously evoking more auditory nerve fibers and later CANS [8]. Therefore, this characteristic of the chirp stimulus, with its opposite frequency order, is beneficial in improving the total number of neural activities in the short-latency responses, leading to enhanced amplitude size with clear morphologies and more accurately balanced neural activities across the frequency range [8,11]. Furthermore, one of the reported limitations of BIC studies is the small peak-to-peak amplitude of wave V in ABR to observe the amount of attenuation correctly [6].
The first aim of the current study was to examine BIC differences using multiple stimuli, including 500 Hz tone-burst (hereafter referred to simply as TB500), click, and chirp. The wave V peak-to-peak amplitude was analyzed to ensure the stimulus effect on BIC. The ABR results elicited by the chirp stimulus were expected to present enhanced peak-to-peak amplitude in wave V for both monaural and binaural conditions, resulting in improved BIC outcomes.
Secondly, this study compared the frequency effects of ITD and ILD to the BIC. According to the duplex theory, ITDs are used for low-frequency localization, and ILDs for high-frequency localization in the horizontal plane [12]. However, these frequency-specific uses of ITD and ILD cues do not accurately translate to experiments conducted using headphones [13]. Complex yet distinct neural networks responsible for processing frequency-specific ITD and ILD cues may be represented through neurophysiological BICs.
Lastly, we hypothesized that individuals with a larger BIC would perform better in binaural hearing ability based on psychoacoustical measurements in a normal listener group. Some studies reported evidence of a correlation between the amplitude of BICs and behavioral ITD and ILD [14,15]. We believed that listeners in the normal hearing category might have varying degrees of binaural hearing ability because of the comprehensive involvement of the peripheral and central auditory systems.
Subjects and Methods
Participants
A total of 16 adults, consisting of 4 males and 12 females with normal hearing, aged between 22 and 25 years (mean: 22.93, standard deviation [SD]: 0.85), participated in this study. Prior to the experiments, all participants completed demographic questionnaires, revealing no history of hearing deficits. Pure-tone audiometry and tympanometry were performed to confirm hearing sensitivity, with all participants exhibiting thresholds within the normal range (<25 dB HL) from 250 Hz to 8,000 Hz. The group mean pure-tone average (PTA), calculated as the mean threshold across 500, 1,000, 2,000, and 4,000 Hz, was 4.37 dB (SD=4.78), with all participants demonstrating type A tympanograms. The experimental procedures were approved by the Institutional Review Board of Tongmyong University (TUIRB-2022-003).
ABR recording and data management procedure
ABR was recorded in an acoustically and electrically shielded soundproof booth. The Eclipse equipment (Interacoustics Ltd.), a conventional clinical AEP device, was utilized for ABR recording, employing Etymotic Research ER-2 insert earphones as the stimulus transducer. Ambu Neuroline 720 surface electrodes (Ambu Inc.) were prepared to construct the electrode montage. A two-channel vertical electrode montage was selected, with the non-inverting electrode on the high forehead (Fz), the inverting electrode on bilateral mastoids, and the ground electrode on the low forehead (Fpz), to record the evoked potentials. The impedance level was below 3 kΩ in all conditions. Three transient onset stimuli—CE chirp LS, click, and TB500—were presented with the rarefaction polarity at 65 dB nHL. A Blackman window function was applied to TB500, featuring 5 sine waves in 10 ms with 4 ms rise and fall times and a 2 ms plateau. The stimulus rate was 32.1 Hz, and 3,000 sweeps were presented in each condition. Bayesian weighting was employed for more stable recording in the presence of undesirable noises or artifacts. The high and low pass filters were set at 100 Hz and 3,000 Hz, respectively, and the artifact rejection was ±40 μV. Three conditions (monoaural right, monoaural left, and binaural stimulation) were recorded to observe the binaural and monaural electrophysiological responses. Two trials were collected in all conditions, and the acquisition orders were randomized to eliminate potential bias. Participants were comfortably relaxed on a bed with a neck pillow and instructed to close their eyes and to sleep for the experimental period if possible. Peak-to-peak amplitudes of wave V were determined offline by three independent audiology professionals with a minimum of 3 years of clinical experience in ABR recording. The BIC was computed by subtracting the amplitude of wave V obtained with binaural stimulation from that obtained by the summed response from the right and left monoaural stimulation using Microsoft Excel 365.
Behavioral ITD and ILD test procedure
Behavioral ITD and ILD tests were conducted using the PSYCHOACOUSTICS software [16]. The program ran on an LG ultra PC laptop, generating acoustic signals delivered through high-quality Sennheiser HD600 headphones. The acoustic stimuli transmitted by the headphones were calibrated according to the software guidelines [16].
All participants were seated in the center of a sound booth wearing headphones. We employed a two-alternative forced-choice (2AFC) paradigm, where two different stimuli were presented in a random order, and participants had to select dichotic condition as their response. Participants were instructed to control the software and select the stimulus that sounded like it was moving within the head. When presented over headphones, ITD for high-frequency signals above 1,500 Hz were perceived as being at the center of the head, regardless of the degree of ITDs [17]. Unlike ITD resolution, ILDs are consistently detectable across all frequencies exhibiting relatively equal [18] or lager [19] sensitivity across frequencies. In this regard, the ILD stimuli consisted of frequencies at 500, 2,000, and 4,000 Hz in this study. The ITD stimuli comprised sinusoids at 500 Hz and 2,000 Hz, with the exclusion of the high frequency 4,000 Hz due to the limitation in ITD delivery.
In each stimulus condition, four sinusoid tones were presented in series. One stimulus condition had 0 ITD or 0 ILD, creating the perception of sound at the center of the head. This 0 ITD or ILD stimulus was designed by presenting all four sinusoid tones simultaneously (ITD) or at the same level (ILD) in both ears (diotic). The other stimulus condition carries ITD or ILD alternating between 0 and a positive non-zero ITD or ILD values based on the participant’s decision. In the alternating stimulus, the first and third tones were diotic (0 ITD or 0 ILD), while the second and fourth tones had varying ITD or ILD, creating the perception of sound moving between the two ears. The inter-tone interval was set at 100 ms, with a 200 ms gap between the first and second stimuli. Each sinusoidal tone consisted of a duration of 400 ms with a 20 ms rise/fall time. The stimulus intensity was set at 65 dB SPL in all conditions. Each test condition was repeated twice to confirm repeatability. The initial ITD was set to 0.2 ms, and ILD was set to 3 dB SPL. ITD and ILD thresholds were estimated based on a 2-down 1-up procedure in the software. After two consecutive correct responses, the ITD or ILD decreased; after one incorrect response, it increased. This adaptive procedure maintained an ITD or ILD level corresponding to a 70.7% correct performance threshold [20]. The signal magnitude of ITD or ILD for the next trial is determined by multiplying (for incorrect response) or dividing (for incorrect response) the value for the current trial by a certain factor.
Statistical analysis
We used SPSS version 29.0 (IBM Corp.) for statistical analyses. Paired-sample t-tests compared wave V amplitudes in the summed monaural response (R+L) to binaural stimulation to assess BIC presence for all stimulus types. Repeated measures ANOVAs, with Bonferroni adjustments for pairwise comparisons, were used to examine potential differences in BIC across stimulus types. A paired-sample t-test was used to compare ITD thresholds between 500 Hz and 2,000 Hz, and repeated measures ANOVA analyzed ILD threshold differences at 500 Hz, 2,000 Hz, and 4,000 Hz. Pearson correlation coefficients assessed associations between ITD and ILD thresholds, and between electrophysiological BIC and behavioral ITD/ILD outcomes.
Results
Presence of BIC
Fig. 1 presents the derived waveforms for all stimulus conditions in subject 3. The original waveforms recorded by the AEP equipment were reconstructed offline by extracting the μV values at 0.2 ms intervals and interpolating the points for visual clarity. Each waveform represents the average of two recordings per stimulus condition. The right, left, summed monaural, binaural responses, and BIC waveform are clearly shown. The DV in the BIC waveform reflects the amplitude difference of wave V, calculated by subtracting the binaural response from the sum of the two monaural waves. The DN1 represents the successive negative peak following DV in the BIC waveform. In this visual representation, chirps produced robust wave Vs and DVs (BIC) compared to other stimuli. Consistent with prior studies [21,22], we designated the BIC as DV. Notably, chirp-evoked waveforms exhibited the earliest latencies, while TB500 evoked the longest latencies.
Illustrative examples of ABR waveforms derived from one participant (subject 3). The waveforms are arranged side by side based on stimulus types (chirp, click, and TB500). The waveforms are displayed from top to bottom according to stimulus side: red (right monoarual), blue (left monoaural), purple (summed monoaural), black (binaural), and grey (computed BIC: binaural minus summed monoaural). ABR, auditory brainstem response; BIC, binaural interaction component.
In our study, it was observed that across all stimulus conditions, each participant demonstrated positive values in the BIC calculation, indicating the prevalence of BIC presence. However, an exception was noted in subject 3, who exhibited a negative value (-0.01 μV) in the TB500 condition. In other words, all 16 participants demonstrated evidence of BICs for both chirp and click stimuli. Fifteen out of 16 participants exhibited evidence of BICs for TB500. For the chirp stimulus, the group mean amplitude of right and left summed wave V was 1.12 μV (SD=0.27), while binaural wave V amplitude was 0.70 μV (SD=0.20). For the click stimulus, the group mean amplitude of right and left summed wave V was 0.73 μV (SD= 0.27), with a binaural wave V amplitude of 0.50 μV (SD=0.15). In the case of the TB500 stimulus, the group mean amplitude of right and left summed wave V was 0.59 μV (SD=0.15), and the binaural wave V amplitude was 0.47 μV (SD=0.16). To confirm the statistical significance of the identified BIC, comparisons were made between the amplitude of wave V in the summed monoaural response (R+L) and that of binaural stimulation. Paired-sample t-tests were conducted based on the type of stimulus. Notably, the amplitudes of wave V elicited by summed monaural stimulation significantly exceeded those produced by binaural stimulation (p<0.001), a consistent trend observed across all stimulus types. Comprehensive statistical details, including mean, standard deviation, and t-test results, are presented in Table 1 for reference.
Difference between amplitudes of wave V for all pairs of binaural and summed monoarual conditions across the stimuli
Comparison of BIC for three stimuli
Fig. 2 illustrates the mean amplitude of BIC across three stimulus conditions. Specifically, the mean amplitude was observed to be highest for the chirp stimulus (mean=0.42 μV, SD=0.20), followed by the click stimulus (mean=0.25 μV, SD=0.16), and lowest for TB500 (mean=0.10 μV, SD=0.06). A repeated measures ANOVA was conducted to evaluate the differences in mean amplitude across the stimuli, representing a significant result (F(2, 30)=17.815, p<0.001). Consequently, post hoc analysis, utilizing Bonferroni adjustment, not only revealed significant differences between the chirp stimulus and both the click (p<0.001) and TB500 (p<0.001), but also indicated significantly higher amplitudes for the click stimulus compared to TB500 (p<0.05).
Mean amplitude of BIC for three stimulus conditions. Error bars represent standard errors. *p<0.05; ***p<0.001.
Behavioral ITD and ILD results
Figs. 3 and 4 represent the group mean thresholds for ITD and ILD in the psychoacoustic tests, respectively. The group mean for ITD at 500 Hz was 0.09 ms (SD=0.05) and 0.14 ms (SD=0.04) for ITD at 2,000 Hz. A paired-sample t-test was conducted based on the thresholds of different frequencies. It was found that the ITD threshold at 2,000 Hz was significantly higher than ITD at 500 Hz (t(15)=-3.535, p<0.05). For ILD, a statistically systematic pattern of ILD results was not identified across three frequency stimuli. The group mean ILD at 500 Hz was 1.64 dB (SD=0.82), ILD at 2,000 Hz was 1.98 dB (SD=0.58), and ILD at 4,000 Hz was 2.05 dB (SD=0.66). There was no main effect according to the stimulus frequencies for ILD (F(2,30)=2.872, p=0.072) according to the repeated measures ANOVA.
Group mean behavioral ITD results for 500 Hz and 2,000 Hz signals. Error bars represent standard errors. *p<0.05. ITD, interaural time difference.
Group mean behavioral ILD results for 500, 2,000, and 4,000 Hz signals. ILD, interaural level difference.
As indicators of high-level auditory processing, both ITD and ILD represent psychoacoustical behavioral performances that may be interconnected. To explore this potential relationship, a Pearson correlation analysis was conducted to examine the association between ITD and ILD across different frequencies. Results are presented in Table 2, which displays the correlation coefficient and corresponding p-values for each pair. Notably, only a subset of ILD pairs demonstrated significant correlations. Specifically, a robust positive correlation was observed between ILD at 500 Hz and ILD at 2,000 Hz (p<0.05), as well as between ILD at 2,000 Hz and ILD at 4,000 Hz (p<0.01). No significant relationship was identified among other pairs.
Correlation matrix of behavioral ITD and ILD results
Association between electrophysiological and behavioral outcomes
Fig. 5 presents scatter plots illustrating the correlation between ITD and binaural BIC across all individuals. One of the primary objectives of this study is to ascertain whether electrophysiological BIC measures align with behavioral outcomes of lateralization. To explore this, Pearson correlation analysis was conducted to evaluate the strength and significance of the relationship between BIC elicited by three different stimuli and two ITD conditions. The analysis revealed a significant correlation between chirp BIC and ITD at 500 Hz (R=-0.545, p<0.05). However, no significant correlations were observed between other pairs of BIC stimuli and ITD frequencies.
Scatter plots of BIC amplitudes as a function of ITD thresholds across stimulus types. BIC, binaural interaction component; ITD, interaural time difference. *p<0.05, statistically significant.
The relationship between electrophysiological measures of BIC and behavioral ILDs across different stimulus conditions was also investigated. Fig. 6 illustrates the p-value of the correlation between BICs and behavioral ILD thresholds, with the resulting correlation coefficients presented for all pairs. Notably, at 500 Hz, ILD displayed a broad range of ITD thresholds, with some data points falling below 1 ms. The greatest amplitude of chirp BIC can be also represented followed by click BIC and TB500 BIC. Pearson correlation analysis revealed that, except for chirp BIC and ILD at 500 Hz (R=-0.533, p<0.05), no significant correlations were found between BICs and ILDs.
Scatter plots of BIC amplitudes as a function of ILD thresholds across stimulus types. BIC, binaural interaction component; ILD, interaural level difference. *p<0.05, statistically significant.
Discussion
The studies of click-evoked BIC have been reported to be detectable in the majority of individuals with normal hearing [1,22,23]. However, a skeptical viewpoint also existed in that the BIC might not be universally present in all individuals with normal hearing. In our study, the BIC was measured in most participants regardless of stimulus type. All 16 participants showed evidence of BICs for both chirp and click stimuli (100%). Fifteen out of 16 participants exhibited evidence of BICs for TB500 (93.75%). Although one participant showed a negative amplitude from the BIC equation at TB500 condition, the amplitude size was negligibly small (-0.01 μV). This high rate of BIC observation satisfies one of the prerequisites for the BIC to be applicable in clinical settings.
Several studies have investigated the relationship between BIC and hearing loss. Hazavei, et al. [24] found that adults with moderate sensorineural hearing loss exhibited smaller BIC amplitudes than individuals with normal hearing, likely due to temporal processing deficits in contralateral inhibitory neurons within the MSO and LSO. Conversely, Adarsh, et al. [25] reported no significant differences in BIC amplitudes between individuals with normal hearing and those with hearing impairment. The presence of high inter- and intra-individual variability in ABRs could potentially explain these discrepancies. To establish the BIC as a clinical biomarker with global consensus, further research is needed to examine systematic patterns of BIC in relation to other hearing-related variables such as the degree of hearing loss.
It is well-known that greater monaural ABR responses can be recorded by chirp stimuli in comparison to conventional clicks in general ABR recordings [8]. The present study aimed to compare the amplitude differences of wave V and the BIC elicited by chirp, click, and TB500 stimuli. Our investigation revealed that the amplitudes of wave V and BIC were notably higher for chirp stimuli compared to click and TB500 stimuli. This finding aligns with previous research by Riedel and Kollmeier [6], where BICs were recorded from 10 individuals with normal hearing for levels ranging from 10 to 60 dB HL. Similar to our study, they also reported a larger amplitude of wave V and BIC for chirp over click stimulus. The enhanced synchronization elicited by lower frequency area in cochlea with chirp stimulus is attributed to the combined contributions of both low- and high-frequency neurons, resulting in an amplification of wave V and BIC.
Some studies have reported that at higher stimulus levels (over 60 dB nHL), chirp stimuli do not consistently produce larger wave V amplitudes compared to clicks [26,27]. Similarly, Riedel and Kollmeier [6] found that while BICs were observed up to 60 dB nHL, the larger chirp-evoked responses reached saturation, becoming closer in amplitude to those evoked by clicks at higher levels. They suggested that the benefits of chirp stimuli over clicks are mainly seen at low to intermediate levels (around 40 dB nHL). In our study, the intermediate level of 65 dB nHL yielded a significant advantage in BIC amplitude. This suggests that chirp-induced BICs at higher stimulus levels may have potential for individuals requiring stronger stimulation, such as those with significant hearing loss. However, if the advantage of chirp stimuli is limited to lower and intermediate levels, patients with considerable hearing loss, who need higher stimulation levels to elicit ABR, may not benefit. Most BIC studies on individuals with hearing loss [25,28] use high-level click stimuli to elicit robust ABRs and obtain BICs. The relative superiority of chirp stimuli, which have a level-dependent limitation, compared to clicks, which produce smaller amplitudes, requires further investigation in future studies.
The aim of the current study is to examine whether behavioral ITD and ILD are associated with electrophysiological BIC for three types of stimuli. Since ABR in response to clicks mainly reflects high-frequency neural activity, it was inferred that click-evoked BICs primarily represent ILD function, while TB500-evoked BICs may reflect ITD function. However, our results showed that only ITD and ILD thresholds at TB500 were significantly correlated with chirp-evoked BIC amplitudes, with strong negative correlations (r<-0.5), indicating that higher BICs are associated with lower detection thresholds. The significant correlation of low-frequency (500 Hz) behavioral thresholds with chirp BICs may be due to the greater amplitude of chirp-evoked BICs and the inclusion of low frequencies in chirp stimuli. The chirp stimulus, which evokes broad neural synchrony, led to consistently high BIC amplitudes across participants. Similarly, Biagio-de Jager, et al. [29] found that LS CE chirp ABR thresholds were closer to low-frequency thresholds compared to click ABR thresholds.
In summary, chirp stimuli evoked the highest wave V and BIC amplitudes compared to clicks and 500 Hz tone bursts. Notably, chirp-evoked BICs correlated significantly with low-frequency ITD and ILD thresholds, suggesting their potential for assessing low-frequency binaural hearing. The small sample size in this study may, however, limit the interpretability of the results. Required sample sizes, as calculated using G*Power 3.1, ranged from 10 to 27, depending on the statistical test. Future research incorporating larger sample sizes and additional analysis of BIC latencies will allow for more robust conclusions.
Notes
Conflicts of Interest
The authors have no financial conflicts of interest.
Author Contributions
Conceptualization: Won So, Sungmin Lee. Data curation: Sungmin Lee. Formal analysis: Won So, Sungmin Lee. Funding acquisition: Sungmin Lee. Investigation: Won So, Sungmin Lee. Methodology: Won So, Sungmin Lee. Project administration: Sungmin Lee. Resources: Sungmin Lee. Software: Sungmin Lee. Supervision: Sungmin Lee. Validation: Won So, Sungmin Lee. Visualization: Won So, Sungmin Lee. Writing—original draft: Won So, Sungmin Lee. Writing—review & editing: Won So, Sungmin Lee. Approval of final manuscript: Won So, Sungmin Lee.
Funding Statement
This work was supported by the National Research Foundation of Korea (NRF) grant funded by the Korea government (MSIT) (NRF-2022R1G1A1011727).
Acknowledgments
We sincerely thank all participants for their valuable time and effort in contributing to this study.