Masking Cervical Vestibular Evoked Myogenic Potentials Elicited by Vertical-Axis Vibrations through Speech Noises or Random Interstimulus-Interval Tone-Bursts

doi:10.21203/rs.3.rs-4130464/v1

Download PDF

Article

Masking Cervical Vestibular Evoked Myogenic Potentials Elicited by Vertical-Axis Vibrations through Speech Noises or Random Interstimulus-Interval Tone-Bursts

https://doi.org/10.21203/rs.3.rs-4130464/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Vestibular evoked myogenic potential (VEMP) can be elicited using bone-conduction vibration (BCV) and air-conduction sound (ACS), with BCV VEMP conventionally linked to bilateral vestibular pathways. We employed a new method to obscure BCV VEMP using acoustic maskings, aiming to contribute to the possibility of unilateral BCV VEMP testings. Twenty healthy adults (20–37 years, 10 males10 females) were enrolled. The vertical-axis vibrations (VAVs) of 500-Hz short tone burst (STB500) and 750-Hz short tone burst (STB750) were used to induce cervical VEMP through a Mini-Shaker (model 4810, Bruel & Kjaer) placed at vertex without acoustic masking (NOM), with 100-dBSPL speech noise masking (SNM), or with random interstimulus-interval tone burst (rISITB) were applied binaurally during VEMP testing. While response rates of STB500 were relatively less affected by SNM or rISITB (92.5% for NOM, 85.0% for SNM, and 75.0% for ISITB), response rates of STB750 were significantly reduced from 90.0% (NOM) to 17.5% (SNM) and 45.0% (rISITB) (p < 0.001, Fisher’s exact test). The response amplitude and p13 latency of STB750 were also significantly different from those of STB500 (p < 0.01, two-way repeated measures ANOVA). VAVs of STB750 elicited a >90% response rate of cervical VEMP and had 80% diminish of response rate by SNM. SNM demonstrated superior masking efficacy to rISITB. Although further research is warranted for possibilities of clinical application, our results indicate the methodology here provides potential of conducting VEMP tests on an individual ear and/or a specific organ using BCV VEMP.

Biological sciences/Neuroscience/Auditory system

Biological sciences/Neuroscience/Auditory system/Midbrain

cervical vestibular evoked myogenic potential

bone-conduction vibration

speech noise

masking

random interstimulus-interval tone burst

The vestibular evoked myogenic potential (VEMP) is currently a widely used balance assessment tool in clinical settings. It often employs brief auditory stimuli, such as 500 Hz tone bursts and clicks¹, to evaluate the integrity of the neural pathways associated with the otolithic organs. This assists in the pathological diagnosis of both peripheral and central vestibular nerve disorders^2,3. However, auditory stimuli may not effectively elicit responses or may elicit atypical responses in patients with abnormalities in middle ear function^4–6. In order to compensate this problem, bone-conducted vibrations (BCVs) are used to elicit VEMP responses, and the responses are referred to as bone-conducted vibration vestibular evoked myogenic potential (BCV VEMP).

To elicit a BCV VEMP response, the vibratory stimuli are usually generated using a transducer such as B81⁷, B250⁸, Mini-Shaker^9,10, or hammer tapping¹¹. The energy of BCV circumvents the middle ear, directly transferring into the peripheral vestibular organs to induce the VEMP response. Consequently, BCV VEMP facilitates obtaining information about integrity of the vestibular pathway in patients with middle ear disorders⁶. The amplitudes of BCV VEMPs commonly exceed those elicited by acoustic stimuli. Thus, BCV VEMP confers supplementary clinical diagnostic value, especially when a middle ear disorder is evident¹².

Based on the previous research¹³, the BCVs in the frequency range of 100–800 Hz demonstrates distinct cervical VEMP response (cVEMP) sensitivity. The optimal stimulation frequency was between 200 Hz and 400 Hz, and the BCVs with a frequency exceeding 1000 Hz failed to elicit VEMP responses¹³. Moreover, 500-Hz short tone bursts are the most commonly used stimuli of BCV to elicit cVEMP¹⁰. To elicit stable cVEMP responses, the vibratory energy should be maintained at 135–140 dB peak force level (peFL) or even stronger as the suggestion of the previous research (Govender et al., 2016; Lim et al., 2013).

The primary limitation of BCV VEMP lies in the inherent bilateral nature of BCV stimulations, meaning the responses are elicited concurrently from both vestibular organs including the opposing sides. Presently, the method for unilateral stimulation through BCV is limited. An innovative approach involving the application of masking techniques to selectively inhibit the non-testing side vestibular system may thereby offer a novel avenue for conducting unilateral testing of BCV VEMP. While extant research has demonstrated a diminished response in VEMP elicited by air-conduction sound (ACS) using either white noise masking or BCV masking ^12,14, the effectiveness of masking the BCV response is presently constrained. In this research, we endeavored to employ a novel methodology to obscure the BCV VEMP, with the goal of establishing a foundational understanding that can contribute to the future advancement of unilateral BC cVEMP examinations.

Participants

Twenty healthy volunteers (10 females and 10 males, totaling 40 ears) with the age range between20 and 37years (median 23 years) participated in this research. The otoscopic examinations of all participants were normal, and they all passed the hearing screening tests using pure-tones of 0.25 kHz, 0.5 kHz, 1.0 kHz, 2.0 kHz, and 4.0 kHz at 25-decibelhearing level (dB HL). None of the participants had a medical history of hearing impairment, tinnitus, dizziness, or vertigo. The research procedures were approved by the research ethic committee of Taipei City Hospital, Taipei, Taiwan (TCHIRB-11103021-E-F). The study was performed in accordance with the ethical standards of the Declaration of Helsink. Signed informed consent was obtained from all participants before inclusion of the study.

Production and Calibration of Bone Conduction Vibrations

The stimuli for bone conduction vibrations (BCV) were produced by the IHS SmartEP auditory evoked potential system (Opti-Amp 8002, Intelligent Hearing Systems Corp, Miami, FL, USA) (Fig. 1 “Signal Producer”) and were then amplified using the amplifier (model AVX-5, DAYEN, Taichung, Taiwan) (Fig. 1 “Power Amplifier”) to output the vibrations through the transducer (Mini-shaker type 4810, Hottinger Brüel & Kjaer, Nærum, Denmark) (Fig. 1 “B&K 4810”). The intensity of the vibrations was calibrated. Because of the transient nature of the stimuli, we used peak-to-peak equivalent force level (ppe FL) detected and measured using the 3-axis digital accelerometer (Fig. 1 “ADXL345”, Analog Devices, Wilmington, MA, USA) and the computer system that sampled the vibration signals at the rate of 3k Hz (Fig. 1“MCU Box”, “Laptop Computer”, and “VibroMeter”). The accelerometer was fixed between an 1-KgW weight that mimicked the weight of the vibration transducer (Fig. 1 “1 KgW Weight”) and the transducer during calibrations. The stimulation intensities were also estimated using the artificial mastoid (type 4930, Hottinger Brüel & Kjaer, Nærum, Denmark), the sound levels were read from impulse time weighting (type 2250-S, Hottinger Brüel & Kjaer, Nærum, Denmark) for comparison.

We used bone-conduction vibrations (BCVs) of 500-Hz short tone bursts (STB500) and 750-Hz short tone bursts (STB750) to elicit VEMP responses. The time settings of stimuli were one-cycle rise, two-cycle plateau, and one-cycle fall (1-2-1 cycle), i.e. 2-ms rise, 4-ms plateau, and 2-ms fall for STB500, and 1.33-ms rise, 2.67-ms plateau, and 1.33-ms fall for STB750. Both stimuli were tapered using Blackman window function. The stimulation intensity was calibrated to 155 dB ppe FL using the digital accelerometer, and the intensities obtained from artificial mastoid and sound level meter using impulse time weighting were 140 dB ppeFL for STB500 and 134 dB ppe FL for STB750. Alternating phasing of stimulation was used to reduce response artifacts from the system.

Production of Sound Stimulus for Response Masking

Speech Noise

According to the literature review, the optimal frequency of acoustic stimuli for eliciting ocular VEMP ranges between 300 and 1000 Hz ¹⁵, and the presentation of 95 dB HL white noise to the contralateral ear has been shown to reduce the amplitudes of ACS cVEMP by approximately 41% ¹⁴. Speech noises are bandpass noises with an energy plateau from 250–1000 Hz, a frequency range that optimally evokes ACS VEMP, and 12-dB rollover/octave beyond the band frequency. This type of noise might, therefore, achieve a more effective masking of VEMP response than a broadband noise and also might reduce the discomfort caused by high-frequency high-intensity stimulation. Consequently, speech noises of a calibrated intensity 100-dB SPL (85-dB HL) (SNM) were used to mask the BCV cVEMP response in this research (Fig. 2 “Mask Sound Producer”). The noises were eventually delivered binaurally (Fig. 2. “ER3A”) using insert earphones (ER-3A, Etymotic Research, Elk Grove Village, IL, USA).

Random Inter-Stimulus-Interval Tone Burst

A good VEMP response requires the averaging process of response waveforms. During the response latency, however, an introduction of other stimulus will interfere with the synchrony of signal averaging. The response waveform may be attenuated and even eliminated if the interference continues throughout the acquisition of the response. The acoustic stimulus of 500-Hz short tone burst was reported to be optimal for eliciting ACS cVEMP response ¹⁰. To attempt to mask the response of BCV cVEMP, therefore, we introduced new masking method using the acoustic stimuli of 500-Hz short tone bursts (1-2-1 cycles) at random inter-stimulus-intervals (rISITB). The rISITB were set at a calibrated intensity of 100 dB peak-peak-equivalent SPL (ppeSPL) and was binaurally presented to the subjects using insert earphones (Fig. 2. “ER3A”). Because the response latencies of cVEMP (p1 and n1) evoked by either BCV or ACS were within 25 ms ⁸, the repetitions of rISITB were at an interval range of 32–48 ms, i.e., a rate of 20.8–31.3 Hz.

Vertical-Axis Bone-Conduction Vibration cVEMP

Figure 2 illustrates the example of masking of right-side vertical-axis BCV cVEMP by ACS sound. The vibration transducer was fixed at the vertex of head (Cz) for delivery of vertical-axis bone-conduction vibrations (VAVs).The participant was in sitting position and was instructed to turn his heads 45° to the left side. During recordings, the participants were request to maintain the tension of the neck muscles as steady as possible. The examinee’s hand held the participant’s chin toward the right side to enhance neck muscle contractions. For left-side VAV cVEMP recording, the head was instructed to turn 45° to the right side and the examinee’s hand held the participant’s chin toward the left side.

The electrode placements followed the common montage used for cVEMP recording (Fig. 3) with reference electrodes at the manubrium of the sternum, active electrodes at the mid-portion of sternocleidomastoid muscle, and ground electrode at the forehead (pFz). The electrode impedance was ensured to be ≤ 5 kΩ for each electrode.

Right-side or left-side test first of VAV cVEMP was arranged randomly for all participants. The test of the other side was therefore arranged on a different day, with at least a 24-hour separation and a counterbalanced design. Both STB500 and STB750 were used to evoke VEMP responses and were also arranged randomly for each subject. While acquiring VEMP response using either of the stimuli, three masking conditions including no masking (NOM), SNM, and rISITB were applied to evaluate the effect of masking on the response, and the order of masking conditions was also randomized.

The responses were bilaterally recorded and were repeated once to confirm the reproducibility. A two-minute break between repetitions was mandatory in the testings. The stimulation setting of BCV was based on the recommendations of the literature review (Table 1) ^10,16,17. The response latencies p13 and n23, p13–n23 amplitude, and VEMP asymmetry ratio (VAR) were obtained for statistical analysis.

Table 1

Setting of Bone-Conduction Vibration
	STB500	STB750
Intensity^∗ (dB ppeFL)	140	134
Polarity	Alternating	Alternating
Duration (rise-plateau-fall time in ms)	2-4-2	1.3-2.7-1.3
Stimulation Frequency (Hz)	500	750
Stimulation Rate(Hz)	7.1	7.1
BandpassFilter(Hz)	30–1500	30–1500
Signal Averaging(sweep)	150	150
Repeated testing(per trial)	2	2
Noise Exclusion(µV)	2375	2375
Analysis Window (ms)	64	64
Gain	5k	5k
STB500, a 500-Hz short tone burst with the time setting of 1-cycle rise, 2-cycle plateau, and 1-cycle fall; STB750, a 750-Hz short tone burst with the time setting of 1-cycle 2-cycle plateau 1-cycle fall.
^∗Measured using artificial mastoid and impulse time weighting; ppe FL, peak-peak equivalent force level.

Meta Analysis

There were only few articles used the similar stimulation settings with the current research according to our literature review of BCV cVEMP. The stimulus was usually 500-Hz short tone burst and the bone transducer was fixed at the forehead. Therefore, using meta-analysis, we compared the p13 latency and p13–n23 amplitude evoked by STB500 in NOM condition of the current research with the reviewed articles ^16,18–21.

Statistical Analysis

Fisher's exact tests were employed to compare the response rates between different stimuli and different masking conditions. Two-way repeated measures analysis of variance (RM-ANOVA) and post-hoc pair-wise comparisons of Bonferroni adjustment were used to compare the response latencies, p13–n23 amplitudes between different stimulation frequencies and masking conditions. Paired sample t-tests were used to determine the difference of VAR between the two stimulation frequencies. Independent sample t-tests were used to compare the p13–n23 amplitudes of the participants with and without response in either SNM or rISITB. The statistical analysis was conducted using the software SPSS Statistics (Version 22, SPSS for Windows Inc., Chicago, IL, USA). Values are presented as mean ± standard deviation (mean ± SD) or mean ± standard error (mean ± SE).

Response Waveforms

The response waveforms and their averages of all participants in six settings are illustrated in Fig. 3. The response waveforms of STB500 in NOM condition displayed as typical biphasic waves (Fig. 3A) with a p13 latency of14.4 ms and an23 latency of 20.7 ms. The p13–n23 amplitude of the average waveform was 27.4 µV. Figure 3B shows the responses elicited by STB750 in the non-masking condition. For the average waveform, the latencies of p13 (13.7 ms) and n23 (19.5 ms) were both shorter than STB500, and the p13–n23amplitude (15.3µV) was also shorter than STB500.

Under SNM, the p13–n23 amplitude of the average waveform seemed to reduce to 17.6µV, and the latencies p13 and n23 were 14.5 and 20.3 ms, respectively (Fig. 3C). Moreover, the average waveform of STB750 under SNM condition did not reveal the typical pattern of a VEMP biphasic wave (Fig. 3D).

Under rISITB masking, the p13–n23 amplitude of the waveforms average elicited by STB500 seemed to reduce to 18.1µV, and the latencies for p13 and n23 were 14.6 and 20.8 ms, respectively (Fig. 3E). Meanwhile, the rISITB masking also remarkably decreased the p13–n23 amplitude of the average waveform of STB750 to 7.2 µV (Fig. 3F). The p13 latency was 13.6 ms, and the n23 latency was 18.5 ms.

In short summary, SNM effectively attenuated the p13–n23 amplitudes of both STB500 and STB750, and the responses of STB750 were much more susceptible to the SNM masking. The rISITB also reduced the responses of STB500 and STB750 but to a lesser extent than the SNM masking.

Response Rates

The response rates of VAV cVEMP evoked by both STB500 and STB750 under various masking conditions are listed in Table 2. The response rates were as high as 92.5% and 90.0% respectively for STB500 and STB750 without masking. Although SNM did not significantly reduce the response rate of STB500 (85.0%, STB500 SNM), the response rate of STB750 was remarkably attenuated by SNM to 17.5% (STB750 SNM, p < 0.05, Fisher’s exact test). The rISITB also significantly decreased the response rates of both STB500 and STM750 to 75% and 45.0%, respectively (p < 0.05, Fisher’s exact test). In summary, for STB500, only rISITB significantly reduced the response rate by 18%. For STB750, SNM abolished 72.5% of the responses, while rISITB also significantly reduced the responses nearly by half.

Table 2

Response rates ofZAVcVEMP evoked by STB500 and STB750 under different masking conditions
	STB500	STB750
NOM	92.5%	90.0%^∗†
SNM	85.0%^∗†	17.5%^†
rISITB	75.0%^‡	45.0%^#
∗p < 0.05,compared with STB750SNMconditionusing Fisher’s exact test
^†p < 0.05,compared with STB750 rISITB condition using Fisher’s exact test
^‡p < 0.05,compared with STB500NOM condition using Fisher’s exact test
^#p < 0.05,compared with STB750NOM condition using Fisher’s exact test

Response Latency and Amplitude

With response rates being reduced significantly by the masking of SNM and rITISB, only the responses with a typical biphasic waveform were submitted to latency and amplitude analysis. The numbers of ear with a typical VAV cVEMP response were indicated in Fig. 4for all the six test settings. In Fig. 3, The average waveforms of STB500 reveal a longer latency than STB750, and the statistical results also confirmed that the p13 latencies of all responses elicited by STB500 were shorter than those evoked by STB750 (df = 1, F = 19.31, p < 0.01, two-way RM-ANOVA) (Fig. 4A). Moreover, neither SNM nor rISITB produced a significant effect on p13 latency (df = 2, F = 1.68, p = 0.28, two-way RM-ANOVA) (Fig. 4A).

The VZV cVEMP amplitudes of STB500 were significantly greater than those of STB750 (df = 1, F = 45.3, p < 0.01, two-way RM-ANOVA) (Fig. 4B). Moreover, the analysis of the main effects using two-way RM-ANOVA indicated that the amplitudes were also significantly different among the three masking settings (df = 2, F = 22.7, p < 0.01, two-way RM-ANOVA). Both SNM and rISITB significantly reduced the VEMP amplitudes of STB500 and STB750 (p < 0.01, two-way RM-ANOVA with pos-hoc comparison with Bonferroni adjustment) (Fig. 4B). Specifically, the mean amplitude of STB500 decreased from 55.8 µV to 40.3 µV, representing a 27.8% reduction masked by SNM. Likewise, the mean amplitude of STB750 decreased from 32.1 µV to 24.7 µV, indicating a 39.5% reduction due to SNM. Notably, there was no statistical difference between the response amplitudes masked by SNM and ISITB (p = 0.97, two-way RM-ANOVA with pos-hoc comparison with Bonferroni adjustment).

Meta Analysis

To compare the cVEMP responses of the current study with previous research, we included three studies that used BCVs of 500-Hz short tone burst to evoke cVEMP ^16,18,21. The results, including our data, are depicted as a forest plot in Fig. 5A. The average p13 latency across the research is 14.5 ms with a 95% confident interval of 14.2–14.7ms. There was no significant differenceofp13latency among the four research and the heterogenicity was low (p = 0.92, I² = 0%, meta-analysis).The pair-wise comparisons of the current study with the previous studies by Wang et al. (p = 0.65, meta-analysis t-test) ²¹, Lim et al. (p = 0.50, meta-analysis t-test) ¹⁶, and Dyball et al. (p = 0.60, meta-analysis t-test) ¹⁸ did not reveal a significant difference of p13 latency. However, there was a difference of the current research from the other three researches. While we fixed the transducer at the vertex of the skull, the transducer for delivery of bone vibrations was fixed at forehead in the other three researches.

We also conducted a meta-analysis using the data from three previously published studies with a similar stimulation setting as our research (Fig. 5B) ^19–21. The statistical results indicated a significant difference among the four researches while the heterogeneity was high (p < 0.001, I² = 96.7%, meta-analysis). The p13–n23 amplitudes of current research were significantly less than the other three researches (p < 0.001, meta-analysis t-test).

VEMP Asymmetry Ratio

The asymmetry of bilateral VEMP responses was evaluated by VEMP asymmetry ratio (VAR) that was defined as the difference between bilateral p13–n23 amplitudes divided by the summation of bilateral p13–n23 amplitudes. The VAR of STB500 (10.0% ± 9.0%, mean ± standard deviation) was not significantly different from the VAR of STB750 (12.7% ± 6.7%) (p = 0.42, paired sample t-test) indicating the vibrations of different frequencies did not make response asymmetry in both ears.

Differences in Response to SNM and rISITB

By SNM, the response rates were significantly reduced for both STB500 and STB750. To investigate the response difference between the subjects with and without a positive response in SNM, the responses of STB500 NOM were classified into the group with positive responses in SNM (STB500–SNMp) and a group with negative responses in SNM (STB500–SNMn). The p13–n23 amplitudes of STB500–SNMp group were not significantly different from those of STB500-SNMn group (p = 0.20, independent sample t-test). However, on classification of the responses of STB750 NOM into the group with positive responses in SNM (STB750–SNMp) and a group with negative responses in SNM (STB750–SNMn), the p13–n23 amplitudes of STB750–SNMp group were significantly greater than those of STB750–SNMn group (p < 0.01, independent sample t-test).

The responses of STB500 NOM were otherwise classified into the group with positive responses in rISITB (STB500–rISITBp) and a group with negative responses in rISITB (STB500–rISITBn). The p13–n23 amplitudes of STB500–rISITBp group were significantly less than those of STB500–rISITBn group (p < 0.05, independent sample t-test). Moreover, on categorization of the responses of STB750 NOM into the group with positive responses in rISITB (STB750–rISITBp) and a group with negative responses in SNM (STB750–rISITBn), the p13–n23 amplitudes of STB750–rISITBp group were significantly greater than those of STB750–rISITBp group (p < 0.01, independent sample t-test).

The above results imply the subjects with a positive response in either SNM or rISITB had larger p13–n23 amplitudes of cVEMP response to the BCVs of STB500 and STB750.

In the course of this investigation, cVEMP responses were acquired using VZV of both STB500 and STB750 among a cohort of healthy participants. Concurrently, air-conduction sounds of SNM or rISITB were introduced to obfuscate the cVEMP responses. The findings elucidate a notable reduction in the response rate of cVEMP elicited by VZV STB750, decreasing from 90–17.5% with SNM, and from 90–45% with rISITB. Furthermore, the response rate of STB500 exhibited relative resilience to the influence of both SNM and rISITB. The discerned masking effects of SNM and rISITB on VEMP responses elicited by BCV suggest their potential applicability in foundational scientific pursuits and clinical contexts alike.

Within the domain of VZV cVEMP elicited by STB750, the response rate in the absence of any masking was recorded at approximately 90.0%. The response rate was close to that of STB500 without masking. The result corroborates the discovery documented in the existing literature²², wherein comparable cVEMP assessments were conducted using 500–Hz short tone bursts produced by a mini-shaker positioned at the vertex (Cz) and yielded a corresponding response rate of 90.0%.However, the SNM abolished more than 80% of the positive responses evoked by STB750, and the rISITB also attenuated 55% of the positive responses. The statistically significant disparity of response rate might come from the observation that SNM continuously outputted the masking energy throughout the test duration. In stark contrast, the rISITB that presented at the rate of 25 Hz and a random variability of 20% that resulted in interference bursts manifesting at the duration of 8 ms and the intervals of 32–48 ms. There were gaps of masking energy in rISITB, and the masking effects might therefore be less effective than SNM. Moreover, based on the subjective evaluations of six participants, the discomfort levels of SNM were scored 7–8 out of 10, while those for rISITB were scored 6–7 out of 10. The rISITB appeared to cause a lower level of discomfort than SNM.

Our findings elucidate that the cVEMP responses elicited by STB500 remained unaffected by either SNM or rISITB. In contrast, the response amplitudes of STB750 exhibited a significant reduction in the presence of either SNM or rISITB. Notably, 17.5% of ears still manifested a positive response in SNM. Our results affirm that ears presenting a positive response in SNM also demonstrated a markedly stronger response in NOM condition compared to those lacking a response in SNM. Individual variations in cVEMP responses to BCVs may contribute to the noted disparity in responses to SNM. Further research with a modification in the methodology of acoustic masking is necessary to investigate the effectiveness of complete masking of cVEMP responses evoked by the VZV of STB750, including adjustments of sound energy, sound frequency, and the presentation speed of masking sounds.

In this study, the response rate of STB500 without any masking was 92.5%. This rate aligns with previous cVEMP studies that utilized bone vibrations as stimulations, reporting response rates of 90–100% ^15,19. However, it is noteworthy that the bone vibrator was fixed at the Cz position in our research, differing from the Fz position used in these earlier studies. In a previous study, it was noted that the introduction of 500-Hz BCVs at an intensity of 120 dB pFL resulted in a significant decrease (10–30%) in the response amplitudes of cVEMP induced by air-conduction sounds ¹². Additionally, when white noises of 95-dB HL were introduced to the contralateral ear, there was an approximately 41% reduction in the cVEMP amplitudes evoked with STB500 ¹⁴. While our findings indicated significant attenuations in cVEMP response amplitudes, it is crucial to emphasize the distinction that our stimuli were BCVs and the air-conduction sounds were used for masking of the responses. The above finding simply the bone-vibration energy and sound energy should have interference with each other while evoking the responses of cVEMP, at least at the stimulation frequency of 500 Hz.

In the literature review, the stimuli with a higher stimulation frequency resulted in a shorter response latency ^13,23,24. Moreover, in an oVEMP research that utilized the air-conduction sounds with various rise/fall time values to evoke oVEMP response, the latencies were significantly affected by the rise/fall time of the stimuli ²⁵. A longer rise/fall time resulted in a longer latency, and vice versa. In our study, the rise/fall time of STB750 was 0.67 ms shorter than that of STB500. Our findings indicate that the p13 latencies of VZV STB750 were significantly shorter than VZV STB500 by approximately 0.8 ms. These results align with the aforementioned research and further supported the association between stimulus characteristics and response latencies in VEMP.

Based on the literature review, the response amplitudes of cVEMP induced by BCVs exhibited an association with vibration frequency. Specifically, the cVEMP response amplitudes evoked by vibrations of 200Hz and 400Hz were significantly stronger than those produced by 800–Hz vibrations ¹³. The findings of our current research, which identified a significant predominance of response amplitude for STB500 over STB750, align with the discovery in the aforementioned study. The results suggested that the otolith organs exhibit differential sensitivity to vibrations of various frequencies, leading to differences in the response amplitude of the evoked response.

The anatomical alignment of the macula sacculi approximates parallelism to the vertical axis of the body, while the axis of the macula utriculi is more congruent with the horizontal axis of the body ²⁶.Conventionally, the transducer employed for administering BCVs is positioned at the forehead (Fz) to elicit the VEMP response. The assumption underlying this approach is that the vibrations simultaneously activate both the saccule and utricle, in accordance with their presumed anatomical orientations. However, when the transducer is affixed to the vertex position (Cz), as that undertaken in the present study, the resultant vibrations may preferentially stimulate the saccule over the utricle. Consequently, there is a likelihood that the VEMP responses may reduce in amplitude but do not change in latency. In our investigation, the observed response amplitudes of the STB500 were found to be smaller than those reported in the literature ^19–21 through meta-analysis. These outcomes are consistent with the above-mentioned mechanism.

Both stimulus phase of BCV and stimulation location have major effects on VEMP response especially for oVEMP ²². The initial peak amplitudes of oVEMP demonstrated reciprocal polarities for BCVs of STB500 with compressive and rarefactive phases regardless of stimulation locations ²². The setting of alternating phasing for BCV stimulations might reduce or even cancel the oVEMP response. However, the initial peak of cVEMP presented identical polarities for BCVs of STB500 with compressive and rarefactive phases regardless of stimulation locations. The setting of alternating phasing for BCV stimulations could also elicit a good cVEMP.

The significant characteristics, including reduced response amplitude at higher stimulation frequencies and decreased amplitude with the fixation of the vibration transducer at Cz, indicate that the cVEMP response evoked by VZV STB750 can be effectively suppressed by both SNM and rISITB. This methodology offers the potential for conducting VEMP tests on an individual ear and/or a specific organ (macula) through BCVs.

The placement of the vibrator at the Cz position and a higher stimulus frequency allowed an effective masking of responses as well as a good response rate of over 90% for BCV cVEMP. There were still small portion of individuals who had greater response amplitudes without masking presenting a positive response to speech noise masking. A modification in the methodology of BCV stimulation and/or acoustic masking is required to achieve a complete masking of cVEMP responses. Although further research is warranted to explore these possibilities in greater depth in the future, our results indicate that the methodology here provides the potential of conducting VEMP tests on an individual ear and/or a specific organ (macula) through BCVs.

Data availability

The datasets of response latencies and amplitudes are provided as supplement materials.

Acknowledgements

The research was supported by the grant (No. 11201-62-018) from Taipei City Hospital, Department of Health, Taipei City Government, Taipei, Taiwan.

Author contributions

G.-S. L. study design, contributed to data collection of data analysis, manuscript preparation and revision; W.-T. Dai collected data and data analysis; S.-H. L. contributed to data analysis, manuscript preparation and revision.

Competing interests

No competing interests to declare.

Additional information

Correspondence and requests for materials should be addressed to G.-S. L.

Wu, H. J., Shiao, A. S., Yang, Y. L. & Lee, G. S. Comparison of short tone burst-evoked and click-evoked vestibular myogenic potentials in healthy individuals. J Chin Med Assoc 70, 159–163, doi:10.1016/s1726-4901(09)70350-8 (2007).
Colebatch, J. G. & Halmagyi, G. M. Vestibular evoked potentials in human neck muscles before and after unilateral vestibular deafferentation. Neurology 42, 1635–1636, doi:10.1212/wnl.42.8.1635 (1992).
Murofushi, T., Matsuzaki, M. & Wu, C. H. Short tone burst-evoked myogenic potentials on the sternocleidomastoid muscle: are these potentials also of vestibular origin? Archives of Otolaryngology Head Neck Surgery 125, 660–664, doi:10.1001/archotol.125.6.660 (1999).
Wang, M. C. & Lee, G. S. Vestibular evoked myogenic potentials in middle ear effusion. Acta Otolaryngol 127, 700–704, doi:10.1080/00016480601002070 (2007).
Wang, M. C., Liu, C. Y., Yu, E. C., Wu, H. J. & Lee, G. S. Vestibular evoked myogenic potentials in chronic otitis media before and after surgery. Acta Otolaryngol 129, 1206–1211, doi:10.3109/00016480802620654 (2009).
Zhou, G., Poe, D. & Gopen, Q. Clinical use of vestibular evoked myogenic potentials in the evaluation of patients with air-bone gaps. Otol Neurotol 33, 1368–1374, doi:10.1097/MAO.0b013e31826a542f (2012).
Zhang, Y. et al. B81 Bone Vibrator-Induced Vestibular-Evoked Myogenic Potentials: Normal Values and the Effect of Age. Front Neurol 13, 881682, doi:10.3389/fneur.2022.881682 (2022).
Fredén Jansson, K. J., Håkansson, B., Reinfeldt, S., Persson, A. C. & Eeg-Olofsson, M. Bone Conduction Stimulated VEMP Using the B250 Transducer. Med Devices 14, 225–237, doi:10.2147/mder.S317072 (2021).
Govender, S., Dennis, D. L. & Colebatch, J. G. Vestibular evoked myogenic potentials (VEMPs) evoked by air- and bone-conducted stimuli in vestibular neuritis. Clin Neurophysiol 126, 2004–2013, doi:10.1016/j.clinph.2014.12.029 (2015).
Rosengren, S. M., Colebatch, J. G., Young, A. S., Govender, S. & Welgampola, M. S. Vestibular evoked myogenic potentials in practice: Methods, pitfalls and clinical applications. Clin Neurophysiol Pract 4, 47–68, doi:10.1016/j.cnp.2019.01.005 (2019).
Patterson, J. N., Rodriguez, A. I., Gordon, K. R., Honaker, J. A. & Janky, K. L. Age Effects of Bone Conduction Vibration Vestibular-evoked Myogenic Potentials (VEMPs) Using B81 and Impulse Hammer Stimuli. Ear Hear 42, 1328–1337, doi:10.1097/aud.0000000000001024 (2021).
McNerney, K. M. & Burkard, R. F. The vestibular evoked myogenic potential (VEMP): air- versus bone-conducted stimuli. Ear Hear 32, e6-e15, doi:10.1097/AUD.0b013e3182280299 (2011).
Sheykholeslami, K., Habiby Kermany, M. & Kaga, K. Frequency sensitivity range of the saccule to bone-conducted stimuli measured by vestibular evoked myogenic potentials. Hear Res 160, 58–62, doi:10.1016/s0378-5955(01)00333-1 (2001).
Takegoshi, H. & Murofushi, T. Effect of white noise on vestibular evoked myogenic potentials. Hear Res 176, 59–64, doi:10.1016/s0378-5955(02)00741-4 (2003).
Chihara, Y. et al. Frequency tuning properties of ocular vestibular evoked myogenic potentials. Neuroreport 20, 1491–1495, doi:10.1097/WNR.0b013e3283329b4a (2009).
Lim, L. J., Dennis, D. L., Govender, S. & Colebatch, J. G. Differential effects of duration for ocular and cervical vestibular evoked myogenic potentials evoked by air- and bone-conducted stimuli. Exp Brain Res 224, 437–445, doi:10.1007/s00221-012-3323-1 (2013).
Sheykholeslami, K., Murofushi, T., Kermany, M. H. & Kaga, K. Bone-conducted evoked myogenic potentials from the sternocleidomastoid muscle. Acta Otolaryngol 120, 731–734, doi:10.1080/000164800750000252 (2000).
Dyball, A. C. et al. Bone-conducted vestibular and stretch reflexes in human neck muscles. Exp Brain Res 238, 1237–1248, doi:10.1007/s00221-020-05798-8 (2020).
Greenwalt, N. L. et al. Bone Conduction Vibration Vestibular Evoked Myogenic Potential (VEMP) Testing: Reliability in Children, Adolescents, and Young Adults. Ear Hear 42, 355–363, doi:10.1097/aud.0000000000000925 (2021).
Manzari, L., Burgess, A. M., McGarvie, L. A. & Curthoys, I. S. Ocular and cervical vestibular evoked myogenic potentials to 500 Hz fz bone-conducted vibration in superior semicircular canal dehiscence. Ear Hear 33, 508–520, doi:10.1097/AUD.0b013e3182498c09 (2012).
Wang, S. J., Weng, W. J., Jaw, F. S. & Young, Y. H. Ocular and cervical vestibular-evoked myogenic potentials: a study to determine whether air- or bone-conducted stimuli are optimal. Ear Hear 31, 283–288, doi:10.1097/AUD.0b013e3181bdbac0 (2010).
Govender, S. & Colebatch, J. G. Location and phase effects for ocular and cervical vestibular-evoked myogenic potentials evoked by bone-conducted stimuli at midline skull sites. J Neurophysiol 119, 1045–1056, doi:10.1152/jn.00695.2017 (2018).
Fu, W. et al. Effect of Stimulus Frequency on Air-Conducted Vestibular Evoked Myogenic Potentials. J Int Adv Otol 17, 422–425 (2021).
Janky, K. L. & Shepard, N. Vestibular evoked myogenic potential (VEMP) testing: normative threshold response curves and effects of age. J Am Acad Audiol 20, 514–522, doi:10.3766/jaaa.20.8.6 (2009).
Cheng, Y. L., Wu, H. J. & Lee, G. S. Effects of plateau time and ramp time on ocular vestibular evoked myogenic potentials. J Vestib Res 22, 33–39, doi:10.3233/ves-2011-0437 (2012).
Curthoys, I. S. et al. Otolithic Receptor Mechanisms for Vestibular-Evoked Myogenic Potentials: A Review. Front Neurol 9, 366, doi:10.3389/fneur.2018.00366 (2018).

No competing interests reported.

Download PDF

Version 1

posted

You are reading this latest preprint version

Masking Cervical Vestibular Evoked Myogenic Potentials Elicited by Vertical-Axis Vibrations through Speech Noises or Random Interstimulus-Interval Tone-Bursts

Status:

Version 1

Abstract

Figures

Introduction

Methods

Participants

Production and Calibration of Bone Conduction Vibrations

Production of Sound Stimulus for Response Masking

Speech Noise

Random Inter-Stimulus-Interval Tone Burst

Vertical-Axis Bone-Conduction Vibration cVEMP

Meta Analysis

Statistical Analysis

Results

Response Waveforms

Response Rates

Response Latency and Amplitude

Meta Analysis

VEMP Asymmetry Ratio

Differences in Response to SNM and rISITB

Discussion

Conclusions

Declarations

References

Additional Declarations

Supplementary Files

Status:

Version 1