Chapter 6. Bilateral application should always be considered

6. Bilateral application should always be considered

6.1 Introduction

Binaural hearing refers to hearing with two ears. When listening with two ears instead of one, at least four advantages can be distinguished: 1) binaural loudness summation, 2) use of acoustic head shadow to hear better in noisy places, 3) directional hearing and 4) binaural squelch. In normal hearing listeners binaural hearing is obvious, based on accurate processing of bilateral inputs. That is not necessarily the case for patients using hearing devices; firstly, a short introduction of the four advantages.

1. Binaural loudness summation refers to improved hearing owing to summation of sound as perceived by the two ears: the input perceived by either cochlea is summed, leading to an increase of perceived loudness of 3 dB to 6 dB.

2. Improvement in speech recognition in noise owing to effective use of head shadow. Let us assume that the speech is coming from the front and noise from the left. The speech is heard equally well by either ear, but the noise is not. At the right ear, the noise is perceived attenuated (by acoustic head shadow) compared to the left ear. Therefore, the right ear will have the better speech to noise ratio. Selective listening with the right ear, thus ignoring the left ear, leads to better speech perception.

3. Improved localization of sounds, in the horizontal plane. To identify where sounds come from, the two ears have to work together to detect interaural difference cues caused by the different positions of the two ears with respect to the sound source. These interaural differences comprise interaural loudness differences (ILD; owing to acoustic head shadow) and interaural time differences (ITD, difference in arrival times of the sound at the two ears). Head shadow attenuates primarily high frequency sounds ( > 1500 Hz). As a consequence, high-frequency sounds are perceived louder by the ear nearest to the sound source, creating an ILD. Below 1500 Hz, head shadow causes little attenuation. Instead, ITDs (or, related, interaural differences in phase) are the relevant cue to localize a sound source. Sounds arrive earlier at the ear nearest to the sound source, creating an ITD. ITD is the main cue for sound localization of low-frequency sounds. Owing to the limited distance between the two ears, the ITD varies between 0 ms (sound presented in the front of the listener) to a maximum of 0.7 ms (sound presented at the very left or very right of the listener).

4. Binaural squelch refers to central de-masking. Assume a subject is listening to speech coming from the front and noise coming from the left. The speech in the two ears will be nicely in phase (ITD=0). However, the noise is perceived differently by the two ears, not only because of the head shadow, but also because of ITD, caused by the different travelling times of the noise to the left and right ear. The difference in ITD for speech and for the noise can be used to perceptually separate the speech and noise, referred to as binaural squelch.

Bilateral application of BTEs in patients with bilateral sensorineural hearing loss mostly leads to binaural hearing. However, the outcome might be worse than in normal hearing subjects who are listening with two ears (e.g. concerning directional hearing; Bogaert et al., 2006). The reason is that the two independently operating BTEs might jeopardize the interaural difference cues owing to processing delays and compression amplification. Nevertheless, e.g. Boymans et al. (2008) showed an obvious benefit in a large group of patients with bilateral sensorineural hearing loss using bilateral BTEs, with regard to directional hearing and the effective use of head shadow. 

6.2 Binaural hearing with bone conduction devices

When using the air-conduction route, the two ears can be stimulated independently because they are acoustically well isolated from each other. That is not the case for bone-conduction stimulation. The skull transmits the bone-conducted vibrations quite effectively, with little damping. Therefore, with one BCD, not only the ipsilateral cochlea is stimulated but also the contralateral cochlea. This is referred to as cross hearing. Attenuation of vibrations from one cochlea to the other might vary between -20 dB and +20 dB with a median score of 3 to 5 dB; variation between subjects is significant as well as within subjects, between frequencies (Stenfelt, 2012). In other words, with bone-conduction stimulation, the cochleae are poorly acoustically isolated, however, just sufficiently to be able to detect some interaural differences (Stenfelt, 2005).

Table 6.1 presents data on binaural effects when changing from unilateral to bilateral listening in patients using bone-conduction devices. A summary of published Nijmegen papers is presented on Baha application (bilateral Baha in bilateral conductive/mixed hearing loss and unilateral Baha application in unilateral conductive hearing loss, with a second normal hearing ear). Data are added of a group of normally hearing controls, listening with one versus two ears. This overview has been published elsewhere in more detail (Agterberg et al., 2011), however, the number of patients in that publication was lower. All the patients were evaluated with one and the same protocol, and they were using linear (analogue) Baha devices. Binaural summation and speech in noise were assessed using speech (sentences). Directional hearing was tested with short (1 s) narrow-band noise bursts (Agterberg et al., 2011).

Table 6.1, first row, presents the mean result (and standard deviations) of the four ‘binaural’ factors as measured in normal hearing controls, firstly, listening with one ear blocked by an ear plug and muff, and secondly, while listening with two ears. The difference is presented. Head shadow and binaural squelch have been measured in one experiment and are not easy to separate, so the combined score is included in the Table. The asterisks indicate values that differ significantly from 0. The next two columns show directional hearing results for low frequency sounds (relying on detection of ITDs) and for high frequencies sounds (relying on the detection of ILDs), respectively. Concerning the directional hearing results, statistically significance was assessed by comparing scores obtained in the unilateral listening situation (not shown) and bilateral listening situation. Thus the first row of the table presents norm values for our set-up.

Let us consider bilateral Baha application in bilateral hearing loss. In principle, for hearing devices with a limited MPO, application should be preferably bilateral because summation of the bilateral inputs will lead to approximately 3-6 dB louder perception. That 3-6 dB can compensate to some extend for the negative gain associated with low MPO, as described in Chapters 2 and 3. According to the table, row 2, bilateral Baha application does lead to the expected improvement of approximately 4 dB. This result is in agreement with the literature (see the systematic review by Janssen et al., 2012). Obviously, the inevitable cross stimulation plays a minor role when considering binaural summation.

N Binaural summation(dB) Head shadow & squelch (dB) Directional hearing, 0.5kHz (degrees) Directional hearing, 3 kHz (degrees)
Controls 10 n.a. 4.6 + 1.6* 7 + 7^ 8 + 10^
Bilat CHL, bilat Baha, acquired 17 4.0 + 2.1* 2.5 + 1.8* 26 + 8^ 25 + 8^
Unilat CHL, acquired 13 2.2 + 1.5* 3.0 + 1.7* 16 + 10^ 20 + 12^
Unilat CHL, congenital 10 1.3 + 2.4 1.1 + 1.9 30 + 13 31 + 18

Note. N.a.: not available, CHL: conductive hearing loss.

* significantly different from 0 (p<0.05); ^significantly improved compared to the unilateral listening condition (p<0.05)

Considering row 3, Baha application in acquired unilateral conductive hearing loss, a small but significant binaural summation score was found (row 3), however, not for the unilateral congenital cases using Baha (row 4). Their score was not significantly different from zero. Effective use of head shadow and squelch are combined in the Table. Separate experiments showed that head shadow with the material used is 4.2 dB for normal hearing ears and 2.8 dB for ears stimulated by Baha. The difference is caused by the position of the ‘microphone’ respectively the eardrum. The eardrum profits more from head shadow than the microphone on the externally worn Baha processor.  Comparing these figures with those of column 2 suggests that binaural squelch for normal-hearing subjects is approx. 0.4 dB and for Baha users with acquired bilateral or unilateral hearing loss close to zero, as measured with the present set-up. In other words, the head shadow effect dominates the response. Surprisingly, the patients with congenital unilateral conductive hearing loss didn’t take advantage of head shadow; a non-significant value was measured (row 4). Thus, on the average, head shadow doesn’t help these patients to better understand speech in noise.

Columns 3 and 4 comprise the mean absolute difference in degrees between the perceived location and the real location of the activated loudspeaker. For normal hearing controls, an error of 7-80 was found, for the bilateral Baha users approximately 250 and for the patients with unilateral acquired hearing loss using Baha, approximately 200. These three groups showed significantly improved scores when changing from unilateral to bilateral hearing. Probably, disturbing cross hearing is the reason for the difference in score between the controls and Baha-users.

With regard to the directional hearing abilities of the patients with congenital unilateral hearing loss; they didn’t improve their scores significantly with Baha. On the average, their aided scores were poor, close to their unaided scores. The most likely reason for this observation is that these patients, unilaterally hearing since birth, use unilateral loudness cues and spectral cues more effectively than the other patients and normal hearing controls (Agterberg et al., 2012). In fact, on the average, in this particular group of patients, profit from Baha application was limited (see row 4). Danhauer et al., (2010) published a systematic review of the literature on benefit of Baha application in unilateral congenital conductive hearing loss. In agreement with the presented data, they concluded that ‘audiological measures generally failed to predict patients’ success or satisfaction with the Baha although most patients perceived some benefit’.

In an attempt to understand these results, we decided to study whether or not age at intervention played a role. Figure 6.1 shows the individual summed outcome of three binaural tests (the head shadow test score and the two directional hearing scores) after transformation of the outcomes to z-scores. Results were available of 20 patients with unilateral congenital conductive hearing loss using a Baha for at least 1 year (data published in Kunst et al., 2008). Figure 6.1 shows the summed z-score of these 20 patients as a function of age at implantation. For reference purposes the same procedure was applied to the data of the patients with acquired unilateral conductive hearing loss (Table 6.1, row 3); a mean summed z-score of just over 10 was found. The Figure shows that that value is achieved by just one of the patients with congenital unilateral hearing loss. Obviously, also negative summed z-scores were found, indicating a worse result with Baha than unaided. A significant effect of age at intervention was not found, owing to a large spread in the data points.

5 Slide2

Figure 6.1. Binaural advantage based on the use of head shadow and directional hearing. Summed z-scores are presented as a function of age at intervention. Baha users and non-users are indicated by different symbols (data from Nelissen et al., 2015).

Fitting of a iBCD in patients with unilateral congenital conductive hearing loss to enable binaural hearing is not always successful. Age at implantation doesn’t seem to play a role.

Recently, Nelissen et al. (2015) studied non-use in this group of patients, with a follow-up of more than 5 years. Non-use is not rare; it is indicated in Figure 6.1. A statistically significant association was found between long-term non-use and the short-term (one year) summed binaural advantage score. Non-use was not related to age at intervention (Nelissen et al.). It seems that several of these patients were not able to integrate the new input with that of the normal hearing ear, what is needed for a binaural percept. Indeed, the mean reason to stop was complaints about interfering noises when using the Baha device. Patients with unilateral acquired conductive hearing loss do profit from Baha application (summed z-score of 10 on the average). They seem to profit from previously developed binaural abilities. A similar problem (not being able to integrate the two inputs) has been reported in sequential cochlear implantation in congenitally deaf children and is referred to as the Aural Preference Syndrome (Gordon et al., 2015). Special training to integrate the bilateral inputs is indicated.

A further question is whether or not adult patients with bilateral congenital conductive or mixed hearing loss do profit from fitting bilateral iBCDs. Bosman et al. (2001) reported that the ‘binaural advantage’ as found in their acquired cases (summarized in row 2, Table 6.1) and those of 6 congenital cases were the same. This suggests that early binaural auditory experience is not a prerequisite for effective use in that group; probably the symmetry in hearing thresholds plays a role (in contrast to the subjects with one normal hearing ear and a Baha-aided ear).

A further remark has to be made concerning binaural summation in patients with bilateral conductive hearing loss. Improved hearing is found owing to binaural summation, in the order of 4 dB. This is of major importance. During the initial fitting of two BTEs or two CIs, when switching on both ear-individually fitted devices for the first time, overstimulation might occur owing to binaural summation. Generally, to deal with that, the volume and maximum output of the individual (thus per ear) fitted devices is lowered, and quite often the software does this automatically. For devices with limited MPO, such a lowering of gain and output is principally unwanted and should be avoided. Overstimulation with today’s BCDs is not possible.

In bilateral conductive or mixed hearing loss, devices with limited MPO should be fitted bilaterally to profit from binaural summation (thus without reducing the gain during the fitting procedure after switching on the second device)

At last, so far, the patients in the bilateral Baha studies had rather symmetric bone-conduction thresholds (Janssen et al., 2012). It is not clear what might happen in case of asymmetric bone-conduction thresholds. The theoretical problem is that the better ear might be hindered or even over-stimulated by the loud Baha device near the worse ear. No data are available on this issue.   

So far (end 2015), only binaural data obtained with percutaneous BCDs have been published. Transcutaneous BCDs have also been used bilaterally, however, no data on binaural hearing have been published yet.       

6.3 Binaural hearing with middle ear implants

As far as we know, studies are missing on binaural hearing after bilateral application of middle ear implants like VSB, Otologics MET or Codacs in patients with conductive or mixed hearing loss. Even for pure sensorineural hearing loss, reports are scarce. Three reports were found; Garin et al. (2010) published multi-centre data on 15 patients with sensorineural hearing loss, fitted bilaterally with the VSB. Only binaural summation was measured and a value between 3 to 5 dB could be calculated from their results. This is within the normal range; see Table 6.1. Wolf-Magele et al. (2016) found a lower figure, 1.3 dB (mean of 10 VSB users, 6 with sensorineural hearing loss, 4 with mixed loss). They also reported on the effective use of head shadow (called binaural squelch by the authors). An effect of 1.1 dB was reported what is lower than those reported in Table 6.1 for normal hearing subjects and patients with bilateral conductive hearing loss. Koci et al (2016) performed directional hearing measurements (10 VSB users; 8 of them had sensorineural hearing loss). They found similar results for the aided and unaided condition and obviously better results for bilateral application compared to unilateral application. Similar results have been reported for BTE users (Bogaert et al., 2006).

Concerning binaural advantage of VSB application in unilateral conductive hearing loss, one publication was found (Agterberg et al., 2014). The authors compared patients with unilateral aural atresia fitted with either a Baha (n=4) or with a VSB (n=4). Compared to a BCD, there might be an advantage of using a VSB, namely, cross stimulation with a VSB is absent. An anticipated better result with the cross-stimulation-free VSB than with Baha could not be established.

Reports on binaural hearing with bilateral hearing devices in bilateral conductive and mixed hearing loss or unilateral application in unilateral conductive hearing loss are scarce, although more and more patients are implanted. Binaural measures should be reported much more frequently.   

6.4. Binaural hearing and adaptive sound processing

Today’s iBCDs and VSBs make use of, amongst others, adaptive sound processing like expansion- and/or compression amplification, noise reduction, feedback reduction and adaptive directional microphones. Interaural cues (ILDs and ITDs) might be distorted by independently working digital sound processors. Sound processing times are relatively long (3 to 9? ms) compared to the maximum possible ITD (0.7 ms) and the two devices (or one device and one normal ear) are not synchronized. All these factors deteriorate interaural cues, resulting potentially in poor horizontal localization (Bogaert et al., 2006; Beck & Sockalingam, 2010). In principle, these processing options are not really necessary when fitting patients with good cochlear function (conductive or mixed hearing loss with mild to moderate SNHLc), because the auditory system can still sort out the important information. On the other hand, some of these processing options might be necessary to deal with limitation of the devices such as feedback, the unnatural position of the microphone, noise of the microphone and the limited MPO. Research is needed to assess binaural benefit, especially sound localisation, with these digital sound processors.

It is not evident that patients with conductive hearing loss or mixed hearing loss with a minor to moderate SNHLc do profit from advanced adaptive sound processing. Side effects like impaired spatial hearing might occur. Conclusive research is lacking.

Next chapter >>>

Appendices >>>

References and Abbreviations >>>