• Open Access

Is cartilage conduction classified into air or bone conduction?

Authors


  • This research was supported by a Health and Labour Science Research Grant for the Sensory and Communicative Disorders from the Ministry of Health, Labour and Welfare, Japan. This study was also supported by JSPS KAKENHI Grant Number 23791924 and Adaptable and Seamless Technology Transfer Program through target-driven R&D, JST Grant Number AS251Z00168P. The authors have no other funding, financial relationships, or conflicts of interest to disclose.

Abstract

Objectives/Hypothesis

The aim of this study was to establish the sound transmission characteristics of cartilage conduction proposed by Hosoi (2004), which is available by a vibration signal delivered to the aural cartilage from a transducer.

Study Design

Experimental study.

Method

Eight volunteers with normal hearing participated. Thresholds at frequencies of 0.5, 1, 2, and 4 kHz for air conduction, bone, and cartilage conductions were measured with and without an earplug. The sound pressure levels on the eardrum at the threshold estimated with a Head and Torso Simulator were compared between air and cartilage conductions. The force levels calibrated with an artificial mastoid at the threshold were compared between bone and cartilage conductions.

Results

The difference in the estimated sound pressure levels on the eardrum at the thresholds between air and cartilage conductions were within 10 dB. In contrast, the force levels at the thresholds for cartilage conduction were remarkably lower than those for bone conduction. These findings suggested that sounds were probably transmitted via the eardrum for cartilage conduction. The threshold shifts by an earplug showed no significant difference between bone and cartilage conductions at 0.5 kHz. At 1 and 2 kHz, the threshold-shifts increased significantly in the order of bone, cartilage, and air conductions. These results suggested that airborne sound induced by the vibration of the cartilaginous portion of the ear canal played a significant role in sound transmission for cartilage conduction.

Conclusions

Cartilage conduction has different characteristics from conventional air and bone conductions.

Level of Evidence

N/A. Laryngoscope, 124:1214–1219, 2014

INTRODUCTION

The auditory pathway to the cochlea is primarily classified into air conduction (AC) and bone conduction (BC). In AC, air vibrations are transmitted to the eardrum, which converts the AC signal to mechanical vibrations that are transmitted via the ossicular chain to the cochlea. However, air vibrations are also transmitted to the skull bone to some extent; therefore, they even are audible in the ear with aural atresia. In conventional BC, mechanical vibrations of the skull are induced by a BC transducer placed on the mastoid or forehead bone. However, when sound is presented by a BC transducer, some airborne sounds from the transducer directly reach the ear canal, and are transmitted to the cochlea in the same manner as AC.[1, 2] Thus, every conduction cannot be defined by a single pathway.

Hosoi[3] found that by attaching a transducer to the aural cartilage (the tragus, in particular) a relatively loud sound is audible even with a negligibly small fixation pressure applied to the transducer. This form of the conduction is referred to as cartilage conduction (CC). One of the possible applications of CC is a hearing device for patients who cannot use AC hearing aids owing to aural atresia.[4-6] A BC hearing aid or a bone-anchored hearing aid (BAHA) has been applied as an alternative to AC hearing aids.[7-9] Unfortunately, the transducer of a BC hearing aid has a relatively large mass; a high contact pressure is needed for the device to function properly. Long-term use can also cause skin irritation, long-continued depressions in the skin, and discomfort.[8] For BAHA, surgery is required, and the portion of the implant exposed to open air can induce infection.[10, 11] The use of CC in a hearing aid does not have these disadvantages. Surgery is not required, and the transducer is lightweight with negligibly small contact pressure.

Investigating whether CC is classified into AC, BC, or another, it is important to address which pathways dominate the sound transmission. Generally, AC thresholds are determined by the vibrations transmitted via the eardrum. For BC, although sounds are radiated into the ear canal,[12] it does not contribute to the thresholds in an open ear.[13] The dominant pathway for BC does not involve the eardrum. Although several studies on CC have been reported,[4-6, 14] the contribution of the pathway via the eardrum has not been revealed.

Theoretically speaking, the sound delivered to the aural cartilage can travel to the cochlea via three possible dominant pathways,[5, 6] as shown in Figure 1. In the first pathway, vibrations of the transducer result in stray sound being transmitted to the air surrounding the transducer. The term “stray sound” is used because CC transducer is not designed to produce sound but rather mechanical vibrations to be transmitted to the cartilage. These vibrations, however, also produce sound in the surrounding air, some of which reaches the ear canal and is transmitted to the cochlea via the conventional pathway for AC. This pathway is termed “direct AC.” In the second pathway, vibrations of the aural cartilage are transmitted to the cochlea via the skull bone. This pathway is termed “cartilage BC,” and it does not involve the eardrum. In the third pathway, vibrations of the aural cartilage are transmitted by CC to the cartilaginous portion of the ear canal. These vibrations induce an acoustic signal in the ear canal that is transmitted by AC to the eardrum. This pathway is termed “cartilage AC.”

Figure 1.

Scheme of the three main theoretical components of cartilage conduction. When an earplug is inserted into the ear canal, it blocks the direct air conduction component.

In order to establish the sound transmission characteristics of CC, we estimated the contribution of each of these three pathways for sound reaching the cochlea. This was done by comparing the thresholds for AC, BC, and CC sound—with and without an earplug for blocking stray AC sound via the direct AC pathway.

MATERIALS AND METHODS

Eight volunteers (4 females and 4 males; 28–37 years old) with normal hearing participated. The experimental procedure was approved by the ethics committee of Nara Medical University, and informed consent was obtained from all subjects. Before the experiment, AC and BC thresholds at frequencies of 0.5, 1, 2, and 4 kHz were obtained for each ear, and we confirmed that these thresholds were within 20 dB hearing level. The ear with lower average BC thresholds for the four frequencies was selected in each subject for the study.

In the experiment, thresholds were measured at 0.5, 1, 2, and 4 kHz with and without an earplug for the four conditions: AC, BC, CC, and BC with a CC transducer worn in the same fashion as the CC threshold measurement. The last condition was carried out in order to evaluate whether the occlusion effect occurs or not when the CC transducer was worn. The thresholds were measured twice and averaged. The first threshold measurement was carried out in the following order: AC, BC, BC with the CC transducer, and CC. The second threshold measurement was carried out in the reverse order. For the first measurement, the thresholds were measured at 1, 2, 4, and 0.5 kHz in order. For the second measurement, they were measured in the reverse order. With regard to the masking, narrow band noise was presented to the opposite ear. The intensity of the masking noise was set at the AC threshold level at the nonobjective ear obtained in the preliminary measurement plus 30 dB.

Threshold Measurement

The thresholds were measured using a transformed up-down procedure.[15] A two-alternative forced-choice (2AFC) procedure was employed. The subjects sat in front of a response box in a soundproof room and wore earphones or transducers. The subjects pushed a button to start the measurement. The first trial started 1 second after the button was pushed. For each trial, two lights were turned on for 500 milliseconds in turn. The interval between them was 500 milliseconds. The stimulus duration was 500 milliseconds with rising/falling ramps of 10 milliseconds. The stimulus was presented during either the first or the second lighting. After the second lighting, the subject was required to respond by pressing the button under one of the two lights. After a 1-second interval, the next trial started. The initial intensity was set well above the expected threshold. The stimulus level was increased by 2 dB after either an incorrect response, or a sequence of a correct response followed by an incorrect response on the next trial. The stimulus level was decreased by 2 dB after two consecutive trials with correct responses. A run consisted of a sequence of changes in stimulus level in one direction only; and testing was continued until eight reversals were obtained. The average of the midpoints of every second run (runs 2, 4, 6, and 8) was used as the threshold estimate. The above transformed up-down strategy converges on a signal level corresponding to the probability of a positive response of 0.707.

Instruments

The testing procedure was programmed using RPvdsEX ver. 6.2 (Tucker-Davis Technologies, Gainesville, FL). The stimulus output and response input were processed using a real-time processor (RP2.1, Tucker-Davis Technologies).

The AC and BC stimuli were presented using earphones (AT-02, Rion, Tokyo, Japan) and a bone vibrator (BR-41, Rion), respectively. These earphones and bone vibrator were calibrated in accordance with ISO 389-1 (1998) and 389-3 (1994), respectively.

Figure. 2A shows the devised CC transducer. It was placed on the cavity of the concha (Fig. 2B). Because the cavity is surrounded by aural cartilage such as the tragus, antitragus, and crus of helix, the CC transducer should be able to efficiently vibrate the aural cartilage.

Figure 2.

Cartilage conduction (CC) transducer (A) and wearing it on the cavity of the aural concha (B). The transducer (A) comprises a piezoelectric bimorph and covering material. A ring made of acrylic acid resin is glued to the transducer tip. The outer and inner diameters of the ring are 16 and 8 mm, respectively. Its thickness is 5 mm. The total weight of the transducer is 6 g. The hole of the ring maintains the connection between the ear canal and the outside air and contributes to avoiding the occlusion effect. An advantage of the method of attaching the CC transducer (B) is that it is held in place by a combination of its own weight and the stiffness of the conchal cartilage. The CC transducer weighs 6 g and the fixation force excluding the stiffness of the conchal cartilage was estimated to be approximately 0.06 N, although the fixation force for BC transducers is usually 5.4 N.[20] The force exerted by the stiffness of the conchal cartilage is similarly relatively low.

The ear impression was made from an addition-cured silicone material. Based on the impression, an acrylic earplug was made to fit the ear canal tightly and to be sufficiently deep to the second bend (Fig. 1). Before and after the measurement, we confirmed that the transducer did not touch the earplug.

Compared to AC and BC Thresholds

If the dominant pathway for CC involved the eardrum in the same manner as AC, the sound pressure level (SPL) on the eardrum at the threshold for CC would be identical to that for AC. The SPLs on the eardrum for CC were estimated based on the output level measured with a Head and Torso Simulator (HATS) (Type 4128C; Brüel & Kjær, Nærum, Denmark), and compared with the AC thresholds calibrated with the HATS. The transducer was placed on the simulator in a manner similar to how it would be worn by human subjects. If the dominant pathway for CC did not involve the eardrum in the same manner as BC, the force level calibrated with an artificial mastoid (Type 4930; Brüel & Kjær) at the threshold for CC would be similar to that for BC. For the calibration, the CC transducer had to be pressed against the artificial mastoid. Therefore, the fixation forces for BC and CC transducers were set at 5.4 N.

Statistical Analysis

Statistical analysis was performed using Excel (Microsoft, Redmond, WA). The data were analyzed using a two- or three-way analysis of variance (ANOVA). The Ryan method was used for post-hoc comparisons. The significance level was set at 0.05.

RESULTS

Figure 3A shows the AC thresholds with and without the earplug. A two-way ANOVA revealed statistically significant effects for the earplug (F[1, 7] = 265.39, P < 0.001) and frequency (F[3, 21] = 7.65, P < 0.01). The AC thresholds significantly increased with the earplug at all four frequencies. The thresholds with the earplug at 0.5 and 1 kHz were significantly lower than those at 2 kHz and 4 kHz.

Figure 3.

Thresholds with and without the earplug for air and bone conduction (AC and BC). Parts A and B show the results of AC and BC, respectively. In order to evaluate the occlusion effect of the cartilage conduction (CC) transducer, the BC thresholds were also measured with the CC transducer on the cavity of the concha. These thresholds were also shown in part B. Vertical bars indicate standard deviations.

Figure 3B shows the BC thresholds with and without the earplug. The results with the CC transducer on the cavity of the concha are also illustrated in Figure 3B. A three-way ANOVA revealed statistically significant effects for the earplug (F[1, 7] = 18.69, P < 0.01). However, the frequency and wearing the CC transducer on the cavity of the concha had no significant effect (F[3, 21] = 1.89, P = 0.16 and F[1, 7] = 0.03, P = 0.86, respectively). Although an interaction between the earplug and frequency was recognized (F[3, 21] = 17.10, P < 0.001), no interactions between wearing the CC transducer and the other factors were obtained. These results indicated that wearing the CC transducer had no affect on the BC thresholds. In contrast, a statistically significant effect for the earplug was obtained at 0.5 and 1 kHz.

The SPLs on the eardrum at the thresholds without the earplug were estimated with the HATS for AC and CC (Fig. 4A). A two-way ANOVA revealed statistically significant effects of conduction-difference (F[1, 7] = 6.59, P < 0.05) and frequency (F[3, 21] = 38.36, P < 0.001). The SPLs for CC at 0.5 kHz was significantly larger than that for AC. In contrast, the SPLs for CC at 1 and 2 kHz were significantly lower than those for AC. No difference was obtained at 4 kHz. Figure 4B shows the force levels referring to 1 μN at the thresholds without the earplug for BC and CC. A two-way ANOVA revealed statistically significant effects of conduction-difference (F(1, 7] = 480.91, P < 0.001) and frequency (F[3, 21] = 47.44, P < 0.001). The force levels for CC were significantly lower than those for BC at all frequencies. Figure 4C shows the force levels referring to 1 μN at the thresholds with the earplug for BC and CC. A two-way ANOVA revealed statistically significant effects of conduction-difference (F[1, 7] = 98.32, P < 0.001) and frequency (F[3, 21] = 8.23, P < 0.001). The force levels for CC were significantly lower than those for BC at 0.5, 1, and 2 kHz.

Figure 4.

Comparisons of the output levels at the thresholds between cartilage conduction (CC) and the other two conductions. Part A shows the sound pressure levels on the eardrum at the thresholds without the earplug estimated with a Head and Torso Simulator. Part B and C show the force levels referring to 1 μN at the thresholds without and with the earplug respectively. Vertical bars indicate standard deviations.

Figure 5 shows the increase in threshold with insertion of the earplug for AC, BC, and CC. Upon inserting the earplug, the threshold-increases for AC were larger than those for the other conditions at all frequencies. The BC thresholds showed an improvement at 0.5 and 1 kHz. For CC, the threshold at 0.5 kHz showed an improvement, while the thresholds at 2 and 4 kHz were poorer with the earplug. A two-way ANOVA revealed statistically significant effects of conduction-difference (F[2, 14] = 105.16, P < 0.001) and frequency (F[3, 21] = 98.49, P < 0.001). An interaction between them was also recognized (F[6, 42] = 13.19, P < 0.001). In multiple comparisons, no significant difference between BC and CC was obtained at 0.5 kHz. At 1 and 2 kHz, the threshold-increases were significantly larger for BC, CC, and AC in that order. No significant difference between AC and CC was obtained at 4 kHz. For CC, the threshold-increases were significantly larger as a function of frequency.

Figure 5.

Threshold-increases with the earplug. Vertical bars indicate standard deviations.

DISCUSSION

Blocking the ear canal with the earplug increased the AC thresholds by 20.2 dB to 36.4 dB. The tendency is consistent with that observed in the previous study.[16] For BC, the threshold at a low frequency is decreased by the occlusion effect, which is consistent with previous studies.[8, 17, 18] Between the BC thresholds with and without the CC transducer in the cavity of the concha, no significant differences were obtained. The results demonstrated that the CC transducer did not result in an occlusion effect.

The HATS is equipped with an auricle-shaped appendage. This appendage is softer than the human auricle, and it may not have the same sound transmission characteristics as human aural cartilage. The airborne sounds may show some difference between human ear and HATS. If the cartilage AC dominates the sound transmission, the SPLs estimated using the HATS for CC are unlikely to be an accurate representation of the true SPLs on the eardrum. The current results showed that the SPLs on the eardrum at the thresholds estimated with the HATS for CC differed significantly from those for AC at the frequencies of 0.5, 1, and 2 kHz. The observed differences in the values between AC and CC were within 10 dB. Considering the difference between human ear and HATS, the results suggest that the dominant pathway of CC is probably direct AC or cartilage AC.

Compared to BC, the lower force levels at the thresholds for CC demonstrated that the direct AC and cartilage AC dominated the transmission without the earplug. Furthermore, even with the earplug, they also dominated it at least below 2 kHz.

The direct AC is identical to conventional AC. The threshold-shifts produced by the earplug insertion at 4 kHz showed no difference between AC and CC (Fig. 5), indicating that at this frequency the direct AC pathway is dominant. In contrast, the threshold shifts for BC and CC at 0.5 kHz were consistent with an occlusion effect.[8] No difference between them indicated no contribution of direct AC to the CC threshold at 0.5 kHz. As mentioned above, cartilage BC did not contribute to the CC thresholds. These findings suggest that cartilage AC dominates the sound transmission. The CC threshold shift increased with frequency indicating that there is a corresponding increase in the importance of the cartilage AC pathway with decreasing frequency. Because the cartilage AC plays a significant role in sound transmission, CC has different characteristics from not only BC but also AC.

For the application of CC to hearing aids, the relatively low transmission efficiency at high frequency is disadvantages to BC. In contrast, the smaller size and lower weight of the transducer and the lower fixation pressure is a significant advantage. Considering both the merit and demerit, CC hearing aid is capable of being an alternative device. Furthermore, in the ear with fibrotic aural atresia, previous study showed more efficient transmission owing to the contribution of a fibrotic pathway.[14] In this case, the transmission of CC sound is dominantly mediated by not the skull bone but the fibrotic tissue that connects to the ipsilateral cochlea. Although cross-over stimulation results in reducing the efficacy of binaural hearing for BC,[19] a binaural CC hearing aid can thus maintain the benefits of binaural hearing.

CONCLUSION

In CC, cartilage BC does not dominate the sound transmission, implying that CC is different from BC. Although the eardrum is involved in the dominant pathway, the threshold shifts by the earplug showed the difference in sound transmission between AC and CC. The current findings suggest that the cartilage AC contribute to the sound transmission in CC and its importance increases with decreasing frequency. Considering the characteristic of the transmission, CC has different characteristics from conventional AC or BC.

Acknowledgement

We thank Mr. Takashi Iwakura and Mr. Kyoji Yoshikawa (Rion Co., Ltd.) for development of the cartilage conduction experiment device.

Ancillary