Why Place a Microphone in the Ear Canal?
Laurel Christensen, Ph.D. and Jennifer Schumacher, Au.D.
Introduction and Rationale
In 2003, ReSoundAIRTM was introduced as the first small, behind-the-ear (BTE) hearing aid with a thin tube that could be coupled to the ear using a completely open dome. At this time, approximately 80% of the hearing aids fit in the US were custom, in-the-ear hearing aid styles, as suggested by the sample surveyed in Kochkin (2002). Custom in-the-ear hearing aids were popular because of their relatively small size, but occlusion was a common side effect that kept people who purchased hearing aids from using them (Kochkin, 2000). The combination of the open dome fitting and digital feedback suppression in ReSoundAIR solved the occlusion problem while also providing cosmetic appeal with the thin tube and small size (Nelson, 2005). Today, it is commonplace for these small, thin-tube hearing aids to place the receiver in the ear. This style, that can still be worn completely open, is known as a receiverin- the-ear (RIE) hearing aid. In 2019, RIE hearing aids made up 78.4% of the hearing aids dispensed in the US (Strom, 2020), completely reversing the trend toward fitting custom in-the-ear products.
Unfortunately, while solving one major problem for users of hearing aids, RIE and BTE hearing aids created another one. The microphones on BTE and RIE hearing aids are located above the user’s ear. This means that sounds processed by the hearing aid are not filtered by the user’s pinna and ear canal. This change in the natural location of incoming sounds impacts sound quality and ultimately interferes with a listener’s spatial perception.
To better understand why hearing aid microphone location matters, a review of outer ear acoustics is worthwhile. The pinna and ear canal shape and amplify sound due to their resonant effects. Resonance varies due to the length of the ear canal and the shape and size of the pinna, thus resonant effects are unique to every person. In general, the combined effect of the ear canal and concha resonances results in an approximately 15 dB increase in SPL at the ear drum from 2000 to 5000 Hz (von Békésy, 1960). Hearing healthcare professionals (HCPs) insert a microphone in the ear to measure these resonances as part of our hearing aid fitting because the individual differences matter for an accurate fitting.
Imagine a situation where there are several sounds in the environment, like the living room in your home. There might be a TV on and then some conversation happening. For a listener with normal hearing, it is quite easy to switch attention between the TV and the conversation.
One reason for this is that listeners with normal hearing can interpret the varying spatial relationships between sound sources. Without these spatial cues, listeners feel as if all the sound is “inside their heads” rather than externally located in the environment. Listeners lose the ability to detect depth and details of sound if spatial cues are taken away. Stated simply, a listener loses spatial perception without pinna acoustics. While spatial perception is not required for listeners to experience better audibility with hearing aids, Gatehouse and Noble (2004) have pointed out that it is important that listeners can “locate, identify, attend to, and switch attention between signals so as to maintain communicative competence and a sense of connection with their surroundings” (p. 86). Indeed, MarkeTrak 10 found that the strongest factor driving user satisfaction was “hearing aid performance and sound,” which includes the ability to tell the direction of sound (Picou, 2020).
The pinna and ear canal also impact sound source localization. Due to the different wavelengths of sound and the shape and size of the ear, head and body, frequencies above 1000 Hz are “shadowed.” How much they are shadowed depends on the location of where the sound is coming from, and these changes are our clues to localization. The technical term for this unique shaping is the head-related transfer function (HRTF). No two people have the same HRTF, meaning that every person hears in a way that is unique to them. These cues help us with localizing sound sources, sound source separation, determining what sounds natural to us, and perceiving auditory distance. In fact, hearing via one’s own HRTFs is the only way to truly experience immersive, natural sound. Ask any gamer about the importance of spatial perception and they will talk at length about headphones with special listening features that incorporate processing with HRTFs to create a more realistic experience. This type of audio incorporated into headphones allows a gamer to hear approaching footsteps and exactly the distance and direction they are coming from. It can mean the difference between living and dying “virtually.”
ReSound ONE microphone and receiver in ear (M&RIE)
ReSound has a history of taking inspiration from the natural ways we hear and listen. Therefore, we developed a receiver with a built-in microphone to be placed inside the ear canal in one small wearable module, putting the sound pick-up location where it naturally belongs. The microphone-and-receiver-in ear (M&RIE) is an option on ReSound ONE RIE hearing aids. ReSound ONE features two microphones on the body of the RIE device, while using M&RIE as a third microphone that picks up sound at the entrance of the ear canal (Figure 1). The sound input from M&RIE is shaped by the user’s own unique acoustic cues for individualized sound quality, spatial perception and localization. The M&RIE microphone is active in quiet and moderately complex listening environments, where spatial hearing contributes importantly to the listening experience. In noisier situations, the directional microphones on the hearing aids are activated for an additional signal-to-noise ratio (SNR) advantage.
Current RIE and BTE hearing aids often employ a feature based on the directional microphone system to “recreate” the spatial cues of the pinna. These pinna compensation techniques shift the directivity patterns of the hearing aid microphones, based on estimates of an average adult HRTF for sound in the horizontal plane. M&RIE differs from this technique by preserving the user’s actual acoustic cues in three-dimensional space, so that it is similar to listening with an open ear. Figure 2 shows measurements taken inside the ear canal of a listener as sound is presented 360 degrees around the head in the horizontal plane. The measurements indicate the intensity of the sound at varying frequencies and azimuths for an open ear, a RIE hearing aid using pinna compensation and a M&RIE hearing aid. Note how the color patterns, which represent the sound intensity, more closely match between the open ear and M&RIE plots. While the pinna compensation algorithm has a general pattern similar to the open ear, there is more detail preserved in the M&RIE measurement.
Candidacy
The placement of a microphone in the canal within the receiver module is made possible by digital feedback cancelation. People with hearing losses ranging from mild to severe can benefit from M&RIE; like any hearing fitting, the coupling to the ear canal depends on the balance between need to reduce occlusion, likelihood of feedback, and need for low frequency gain. For this reason, there are two fitting ranges for M&RIE, which are displayed in Figure 3. For users with normal hearing or a mild hearing loss through 1000 Hz (indicated by the light grey range), a closed dome, tulip dome, or micromold should be fitted. The closed dome and tulip dome provide the same degree of openness, with a 10 dB vent effect at 500 Hz. Note that high frequency thresholds for these users should not exceed 70 dB HL, as there will be virtually no attenuation in the feedback path from the receiver back to the M&RIE microphone with an open fitting. For users with moderate-to-severe hearing loss in the low frequencies (indicated by the dark grey range), a power dome or closed micromold should be fitted and in this case, high frequency thresholds can exceed 70 dB HL, as seen on the figure.
Benefit
For users that are candidates, there is substantial evidence that M&RIE provides benefits compared with listening with microphones above or behind the ear in four areas: sound quality, localization, listening effort, and wind noise reduction.
Sound Quality
The first studies on M&RIE tested the idea that customization of pinna cues would lead to better outcomes on measures of sound quality and spatial perception, compared to a pinna compensation algorithm (Groth, 2020). Five normal-hearing participants evaluated overall sound quality and spatial sound quality of M&RIE under headphones using a sound quality evaluation method developed by Legarth et al. (2012). To make this test possible, the sound stimuli for each listening condition were filtered for each individual participant from varying distances and directions. This created a custom set of filters for each of the five listeners’ ears, that could then mimic M&RIE microphone placement and RIE microphone placement above the pinna (pinna compensation). For overall sound quality, they were to listen for clarity, timbre and naturalness. For spatial sound quality, they were to listen for ability to localize sounds, definition of sound, and spaciousness or sense of the room. The stimuli included an office scene, a cafeteria scene and jazz music. Results can be seen in Figure 4. The average overall quality rating and the average overall spatial quality rating for M&RIE was twice as high as for pinna compensation. What is most striking is the lack of variability in the M&RIE rankings versus the pinna compensation. The ratings of pinna compensation across individuals ranged from poor to nearly as good as M&RIE. This variation is an expected finding because when people have very different anatomical characteristics than the average HRTF, the sound delivered via pinna compensation will be less natural and of inferior quality to that picked up at the M&RIE microphone location.
To determine if the results discussed above would hold when participants were fitted with hearing aids and M&RIE, Jespersen et al. (2020) carried out a sound quality test with ten normal-hearing listeners and ten listeners with bilateral mild-to-moderately sloping sensorineural hearing loss. The participants listened to three sound scenes - a cafeteria setting with a target talker, traffic noise and a train station – with three different hearing aid programs: omnidirectional, pinna compensation (ReSound Spatial Sense feature), and M&RIE. The participants were asked to rate sound quality using attributes such as naturalness, clarity and spatial perception in a paired comparison task.
Results from this experiment are shown in Figure 5 for participants with normal hearing (top row) and participants with hearing loss (bottom row). Data across the listening scenarios were combined for a total of four comparisons. M&RIE was the top choice for sound quality in three out of four comparisons. The listeners with normal hearing preferred M&RIE 87% of the time over omnidirectional and 70% of the time over Spatial Sense, which was a statistically significant preference in both cases. The listeners with hearing loss also showed a preference for M&RIE, with M&RIE chosen 70% of the time over omnidirectional and 57% of the time over Spatial Sense, though this difference was only statistically significant in the M&RIE/omnidirectional comparison. At the conclusion of the data collection, comments from the participants in both hearing groups suggested that M&RIE was chosen based on reduced background noise, increased clarity of speech and better spatial perception.
Localization
The ability to locate the source of a sound in an environment is often a difficult task for people with hearing loss, and one that can be degraded by hearing aid use because of the loss of pinna cues (Akeroyd, 2014). However, localization performance has been shown to improve in users fitted with M&RIE, as compared to hearing aids with traditional microphone placement at the top of the device.
Jespersen et al. (2020) conducted an evaluation of localization comparing unaided, omnidirectional, Spatial Sense, and M&RIE conditions. Ten adults with normal hearing and ten adults with bilateral mild to moderate sensorineural hearing loss participated in the study. The listeners were seated in an array of 12 loudspeakers spaced 30 degrees apart and listened to short bursts of white noise. Their task was to identify the loudspeaker from which the sound originated. Localization performance was measured in “front-back error,” which reflects the percent of time confusions between front and back sound locations were made. The results are shown in Figure 6. Note that lower scores indicate less errors and therefore better localization.
Both participant groups performed poorest in omnidirectional mode and best with no hearing aids (open ear). The mean percent of front-back errors in the group with normal hearing was 42% in omnidirectional, 22% using Spatial Sense and 10% using M&RIE, with 0% errors in the unaided condition. The group with hearing loss had mean percent of front-back errors of 47% in omnidirectional, 29% using Spatial Sense, 18% using M&RIE and 10% while unaided. The pattern of performance was similar between the two participant groups, which demonstrates how microphone placement on the top of a hearing aid can alter localization cues for even people with normal hearing. M&RIE significantly improved front-back localization in both groups over omnidirectional mode (p < 0.05). In addition, unaided open ear performance was significantly better than omnidirectional and Spatial Sense modes for both participant groups; however, there was not a significant difference in unaided and M&RIE performance. M&RIE was the only hearing aid condition that allowed users to localize in a way similar to the open ear.
An additional investigation into localization performance was conducted as part of a two-year longitudinal study following twelve adult users fit with ReSound ONE and M&RIE (Jespersen, 2021). Localization using M&RIE was evaluated at the initial fitting (as in the above study) and after four months of wear time. The results can be seen in Figure 7. The front-back error score at the fitting with M&RIE was nearly 20%, which is very similar to the score for participants with hearing loss using M&RIE from Jespersen et al. (2020). After four months of wearing ReSound ONE with M&RIE, frontback localization errors decreased to 12%. This improvement in front-back localization was statistically significant (p < 0.05), suggesting that users can gain benefit from M&RIE following an acclimatization period of use.
Listening effort
Listening effort can be defined as the mental resources deliberately allocated for listening and attending to auditory tasks (Pichora-Fuller et al., 2016). People with hearing loss can expend a lot of effort when listening in noisy or complex situations, which is correlated with increased listener fatigue (Hornsby, 2013). Various laboratory experiments suggest that hearing aids do appear to reduce listening effort, though it is less clear how particular features (e.g., digital noise reduction) may contribute (Hornsby, 2013; Desjardins & Doherty, 2014; Desjardins, 2016). It was hypothesized that M&RIE could reduce listening effort for hearing aid wearers, due to its inclusion of user-specific auditory cues. To test this idea, an investigation was carried out at Hörzentrum (Hearing Center) Oldenburg in Germany (Quilter et al., 2021). Twenty-four experienced hearing aid adult users with bilateral mild-to-moderate hearing loss participated in the study. The participants were fit with ReSound ONE hearing aids. The Adaptive Categorical Listening Effort Scaling procedure (ACALES) (Krueger et al., 2017) was used to measure listening effort in three conditions: unaided, a traditional receiver and M&RIE. ACALES measures subjective listening effort as a function of signal-to-noise ratio (SNR) – as SNR becomes poorer, listening effort tends to increase. On this test, a lower dB SNR score means less listening effort was needed in that condition. Compared to unaided listening, there was a 2.6 dB reduction in listening effort for M&RIE, and a 1.8 dB reduction for the traditional receiver. This reduction in listening effort was statistically significant when compared to unaided (p < 0.05), though not different between the two receiver conditions. However, upon examination of listening effort ratings between the two receivers, M&RIE shows a consistent trend towards better scores, especially in conditions where less listening effort was required (Figure 8). This suggests that M&RIE may provide an advantage over a traditional receiver in less noisy or complex situations that users typically spend most of their time listening.
Wind Noise Reduction
M&RIE provides natural wind noise protection because the microphone is located inside the ear canal where wind does not create as much turbulence over the microphone opening as it does on top of the pinna. The amount of wind noise reduction afforded by M&RIE was measured in a wind tunnel using an RIE hearing aid and M&RIE mounted on a KEMAR (Groth, 2020). The intensity level of wind noise was measured at varying angles around KEMAR, using three wind speeds: 2 meters per second (m/s), 5 m/s and 8 m/s. For reference, a wind speed of 2 m/s corresponds to a light breeze that can rustle leaves, while a wind speed of 8 m/s is a stronger breeze that can cause small trees to sway (University of Maine, School of Marine Sciences). Wind noise reduction was measured by comparing the intensity level of the wind at the omnidirectional microphones on top of the pinna, and with M&RIE located inside the ear opening. Figure 9 displays the average reduction in wind noise across all angles for the M&RIE versus the traditional microphone location on top of the pinna. At 2 m/s, wind noise was reduced by 19 dB by use of M&RIE, reduced by 15 dB in wind at 5 m/s and 14 dB in wind at 8 m/s.
Andersen et al. (2021) evaluated perceptual annoyance with wind noise using RIE hearing aids and M&RIE. Sixteen adults with normal hearing evaluated the sound quality of wind noise with a speed of 5 m/s, prerecorded on an acoustic manikin in a wind tunnel at three azimuths (0 deg., 135 deg., and 270 deg.). They used three hearing aid settings: traditional omnidirectional microphones with digital wind noise reduction (ReSound Wind Guard feature), omnidirectional without Wind Guard and M&RIE (also without Wind Guard). They rated the annoyance level of each listening condition using a 7-point Likert scale, with 1 = no noticeable wind and 7 = extremely annoying. Results for the study are displayed in Figure 10. Wind originating from in front of the listener (0 deg.) showed the greatest annoyance, while wind from behind (270 deg.) was rated as least annoying – this was regardless of microphone location or Wind Guard. The median ratings of annoyance ranged from 4.5 – 7 in omnidirectional mode, from 5-6 in omnidirectional + Wind Guard, and from 2-5 using M&RIE. This decrease in annoyance with wind noise from M&RIE was statistically significant compared to the other programs, regardless of wind direction (p < 0.05).
Conclusion
M&RIE is an innovative concept that gives hearing aid users the benefit of utilizing their own unique pinna cues in a small, cosmetically appealing over-the-ear device. M&RIE has demonstrated a variety of advantages for users – less perceptible wind noise and its accompanying annoyance, improvements in localization, especially after a period of acclimatization, reduced listening effort in noise, and better sound quality ratings related to naturalness, clarity and spatial perception. ■
References
- Andersen, P., Schindwolf, I., & Jespersen, C. (2021). Less wind noise with M&RIE leads to better sound quality. ReSound white paper.
- Akeroyd, M. A. (2014). An overview of the major phenomena of the localization of sound sources by normal-hearing, hearing-impaired, and aided listeners. Trends in Hearing, 18, 1-7.
- von Békésy, G. (1960). Experiments in Hearing. McGraw-Hill.
- Desjardins, J. L., & Doherty, K. A. (2014). The effect of hearing aid noise reduction on listening effort in hearing-impaired adults. Ear and Hearing, 35(6), 600–10.
- Desjardins, J. L. (2016). The effects of hearing aid directional microphone and noise reduction processing on listening effort in older adults with hearing loss. Journal of the American Academy of Audiology, 27(1), 29–41.
- Gatehouse, S., & Noble, W. (2004). The Speech, Spatial and Qualities of Hearing Scale (SSQ). International Journal of Audiology, 43(2), 85–99.
- Groth, J. (2020). An innovative RIE with microphone in the ear lets users “hear with their own ears”. ReSound white paper.
- Hornsby, B. W. (2013). The effects of hearing aid use on listening effort and mental fatigue associated with sustained speech processing demands. Ear and Hearing, 34(5), 523–34.
- Jespersen, C., Kirkwood, B., & Schindwolf, I. (2020). M&RIE receiver preferred for sound quality and localization. ReSound white paper.
- Jespersen, C. (2021). Localization with M&RIE improves with experience. ReSound white paper.
- Kochkin, S. (2000). MarkeTrak V: “Why my hearing aids are in the drawer”: The consumers’ perspective. Hearing Journal, 53(2), 34-42.
- Kochkin, S. (2002). 10-year customer satisfaction trends in the US hearing instrument market. Hearing Review, 9(10), 14, 18-20, 22-25, 46.
- Krueger, M., Schulte, M., Brand, T., & Holube, I. (2017). Development of an adaptive scaling method for subjective listening effort. The Journal of the Acoustical Society of America, 141(6), 4680-93.
- Legarth, S. V., Simonsen, C. S., Dyrlund, O., Bramsloev, L., & Jespersen, C. (2012). Establishing and qualifying a hearing impaired expert listening panel. Poster presentation at International Hearing Aid Research Conference, Lake Tahoe, CA, USA.
- Nelson, J. (2005). Open solutions – Why, how, and when? Proceedings of the 21st Danavox Symposium, Denmark.
- Picou, E. M. (2020). MarkeTrak 10 (MT10) survey results demonstrate high satisfaction with and benefits from hearing aids. Seminars in Hearing, 41(1), 21–36.
- Pichora-Fuller, M. K., Kramer, S. E., Eckert, M. A., Edwards, B., Hornsby, B. W., Humes, L. E., Lemke, U., Lunner, T., Matthen, M., Mackersie, C. L., Naylor, G., Phillips, N. A., Richter, M., Rudner, M., Sommers, M. S., Tremblay, K. L., & Wingfield, A. (2016). Hearing impairment and cognitive energy: The Framework for Understanding Effortful Listening (FUEL). Ear and Hearing, 37 Suppl 1, 5S–27S.
- Quilter, M., Groth, J., & Krueger, M. (2021). ReSound ONE with M&RIE reduces listening effort. ReSound white paper.
- Strom, K. (2020). Hearing aid unit sales increase by 6.5% in 2019. Hearing Review, 27(2), 6,34.
- University of Maine, School of Marine Sciences. (n.d.). Beaufort wind scale. http://gyre.umeoce.maine.edu/data/gomoos/buoy/php/variable_description.php?variable=wind_2_speed
Laurel A. Christensen, Ph.D. is the Chief Audiology Officer of GN ReSound Group. She holds adjunct faculty appointments at Northwestern and Rush Universities, and is a former member of the Executive Board of the American Auditory Society and a member of the Advisory Board for the Au.D. Program at Rush University. Dr. Christensen received her Master’s degree in clinical audiology in 1989 and her Ph.D. in audiology in 1992 both from Indiana University.
Jennifer Schumacher, Au.D., MWC is a Manager of Audiology Communications at GN Hearing. She is an audiologist and medical writer with over a decade of experience in the hearing aid industry. Her primary area of interests are designing hearing aid studies with a focus on clinical outcomes and creating communications for hearing care professionals, patients and their families.