AN UNCONVENTIONAL FITTING APPROACH

A novel and unorthodox idea for allowing hearing aid wearers directional benefit without requiring them to switch is by fitting directional microphones asymmetrically. This approach could also help avoid conflicts between acoustic and auditory scene analysis for devices with automatic switching. The benefit of asymmetric fitting for the hearing instrument wearer would be that there would be no need for manual switching of the microphone mode, and that one of the devices would always be in the preferred microphone mode.

While an asymmetric fitting strategy such as this may seem counterintuitive, there is evidence that it can provide similar benefit to bilateral directionality while circumventing issues with manual and automatic switching. Bentler et al (2004) evaluated five different fitting strategies including binaural omnidirectional, binaural with three different polar responses, and asymmetric with an omnidirectional response on the left ear and a hypercardioid directional on the right. In terms of measured directional benefit, they found no differences between the asymmetric and the bilateral directional fittings.

Both the asymmetric and bilateral directional fittings resulted in significantly better directional benefit than the bilateral omnidirectional fitting. In addition, subjective ratings of sound quality were similar for the asymmetric and bilateral symmetric fittings.

To further investigate subjective aspects of microphone mode preference Whitmer et al (2006) asked listeners to indicate preference for a unilateral fitting, bilateral directional fitting or asymmetrical directional fitting when listening to moderately loud and loud recorded speechin- noise situations. In addition, listeners specified whether their preferences were based on clarity, comfort or some other factor.

They found clear differences in preference for microphone configuration when the results were analyzed in terms of the listeners’ reasons for their preferences. When clarity was the deciding factor, the bilateral directional and asymmetric directional fittings were equally preferred, which is in agreement with the speech-in-noise testing results of Bentler et al (2004).

The advantage of an asymmetric fitting in terms of audibility of sounds coming from different directions is illustrated in Figure 2. The panel on the left represents a listener fit bilaterally with omnidirectional microphones. The orange and blue shaded areas represent the omnidirectional directional pattern of each device, and the dark blue area represents the combined directional characteristics of the two devices. In this case, the speaker in the front as well as those off to the side and behind will be audible, but not necessarily intelligible.

In the middle panel, the listener is fit bilaterally with directional microphones. For this configuration, the speaker in front will be audible and intelligible, but audibility for the other speakers will be reduced. Finally, the panel on the right shows an asymmetric microphone mode fitting. Since the directional benefit for this type of fitting is equivalent to the bilateral directional condition, both audibility and intelligibility are preserved for the speakers at all locations in this example.

Figure 2: Speakers at different locations are audible with a bilateral omnidirectional fitting (left panel), but not necessarily intelligible. Bilateral directional microphone fittings decrease audibility for speakers beside and behind the listener (middle panel). With an asymmetric fitting (right panel), speakers from different directions are audible, and the listener can shift his attention to any one of them.

Asymmetric fitting has also been evaluated in a field investigation. Cord et al (in press) applied an asymmetric strategy in fitting individuals who had not been successful with selectable directional hearing instruments. Like Bentler et al (2004), no differences in directional benefit measured in the laboratory were found for asymmetric versus bilateral directional microphone configurations. However, both of these conditions yielded better performance than when the test condition was bilateral omnidirectional.

For the field evaluation, subjects kept track of the listening situations they encountered and rated their ease of listening in these situations. They were fit for part of the evaluation with a bilateral omnidirectional response and for part with an asymmetric fitting. The listening situations were assigned to one of two groups: those in which an omnidirectional response should be favored and those in which a directional response should be favored. Analysis of 1,612 subject logs of use situations revealed significantly higher ease of listening for the asymmetric condition than the bilateral omnidirectional condition in situations favoring a directional response.

There was no difference in ease of listening ratings between the two conditions for situations favoring an omnidirectional response. These results are compelling evidence that, not only does the directional benefit of an asymmetric fit extend to real-world situations, but also that it can provide directional benefit to previously unsuccessful users of selectable directionality.

THE SOLUTION: NATURAL DIRECTIONALITY

Natural Directionality describes a bilateral fitting strategy which relies on the superiority of the human auditory system’s cognitive signal processing rather than artificial decision-making by the hearing instrument. This unique approach delivers acoustic information differentially to the right and left ears to afford the user directional benefit without working against higher level auditory scene analysis. When fitting Natural Directionality, a “focus” ear and “monitor” ear are carefully determined based on audiometric data.

The focus ear is fit with a new type of directional processing, Focused Directionality, while the monitor ear is fit with an omnidirectional response, allowing the wearer to remain oriented in the environment. The specific inputs to the two ears do not disrupt the wearer’s cognitive analysis of the simultaneous stream of sounds that make up any real-life situation. Thus, the acoustic information made available to the central auditory system enables the listener to efficiently use his sense of hearing to understand the properties of sound-producing events.

An overview of the literature on microphone preferences and asymmetrical microphone fittings supports that the Natural Directionality strategy is the configuration which provides the greatest overall benefit when audibility, speech intelligibility and ease of listening are considered. Figure 3 summarizes this overview. For audibility, any bilateral directional approach will result in the signal of interest sometimes not being heard. Since the signal of interest is off-axis (not in front of the listener) 20% of the time (Walden et al, 2004), this is a frequent scenario for a hearing instrument wearer. Natural Directionality maintains audibility for off-axis signals via the monitor ear.

For intelligibility, it has been demonstrated that an asymmetrical approach such as Natural Directionality provides equivalent directional benefit to bilateral directionality. Manually selectable directionality depends on the user recognizing appropriate situations and switching microphone mode.Finally, ease of listening in actual use has been shown to be significantly better for the Natural Directionality approach than for other microphone configurations in challenging listening situations.

Figure 3: Literature on microphone mode preferences and asymmetrical microphone fittings supports that Natural Directionality offers the most extensive benefit in terms of audibility, speech intelligibility and ease of listening.

The Aventa fitting software incorporates a “Focus-ear Calculator” that automatically suggests which ear should be fit as the focus ear and which one as the monitor ear. The information used to determine the monitor and focus ears is primarily speech recognition scores and, should these not be available, hearing threshold levels are used. Bilaterally hearing-impaired individuals demonstrate interaural differences in SNR loss even when their hearing losses are symmetrical. Furthermore, their binaural performance is determined by the ear with the least SNR loss (Walden & Walden, 2005). Thus, if speech audiometry data is available in NOAH and indicates a significant interaural performance difference, then the better ear will be selected as the focus ear.

The fitter may also enter speech audiometry results directly in the Aventa fitting software. In the event that speech recognition scores are not entered, the better-hearing ear will be selected for asymmetric hearing losses, and the right ear for symmetric losses. For symmetric losses, the right ear is selected as focus ear based on the knowledge that most individuals have a right ear advantage when listening to verbal input (Bryden, 1982).

In terms of signal processing, Natural Directionality incorporates a new directional algorithm which combines an invariable directivity pattern with more sophisticated directional processing. Focused Directionality is a critical element of Natural Directionality and unique to ReSound Azure. While other directional systems can be fit asymmetrically, the ReSound Azure’s distinctive directionality maintains constant focus yet ensures natural balance in volume and sound quality between the two ears. There are three factors which make this possible.

• Firstly, the directional processing accounts for sound reflections from the head and torso to ensure optimum directional characteristics in situ. This ensures that the directivity achieved when the device is worn on the head is equivalent to what can be measured in a laboratory setup.
• Secondly, the low frequency response for the focus ear is equalized with the response for the monitor ear as part of the directional processing. Other directional systems equalize the low frequency roll-off inherent to directional processing by adding gain in the compressor, which has the effect of adding noise. If applied to an asymmetric fitting strategy, the perceptual effect of this traditional approach to low frequency equalization would result in the focus ear sounding noisier than the monitor ear, particularly in quiet situations.
• Finally, proximity effects – another drawback of directional microphones – are eliminated as part of the ReSound Azure’s directional processing. This means that the loud or boomy quality of near-field sounds such as the wearer’s own voice and wind noise is avoided.

Based on the results of Walden et al (2004), directionality has shown to be favored when a speaker is located in front of and fairly near to the listener and background noise is present. However, hearing instrument wearers are only in this type of listening situation about one quarter of their total hearing instrument usage time (Cord et al, in press). The  asymmetric application of ReSound Azure’s focused directionality ensures that one of the hearing instruments will always be in the preferred microphone mode, thus eliminating disruption of the wearer’s auditory scene analysis. In other words, the wearer is provided directional benefit, but is enabled to shift his attention without interference of the hearing aid sound processing.

Figure 4: The Focus ear calculator determines the optimum fitting configuration for Natural Directionality.