Hearing aid application

As operational characteristics are adjusted directly through the graphic interface of the application, this feature allows classifying the hardware-software complex as a programmable hearing aid.[2]

HAA, as a rule, adjusts the level of the surrounding acoustic environment (sound) to the user's hearing characteristics, thereby allowing him to hear better.[3][4][5][6][7][4] Most of HAA works both with wired and bluetooth headsets/headphones.[6][7][4][8][9][10]

Unlike traditional hearing aid, which is a part of system consisting of a hearing aid, wired or wireless interworking interface, computer device and installed software for changes,[2] adjustment of HAA is a procedure within the united hardware-software complex.

Thus, HAA is not inferior, and even surpasses traditional hearing aids in some characteristics (number of channels, sampling frequency etc.).

The basic hearing aid setting is amplification adjustment in accordance with the user's audiogram.

The process of selection of traditional hearing aid starts with audiometry, which is also the first characteristic of sound perception disorder considered during the adjustment of HAA. If the user has already had a special audiologic checkup, this type of adjustment can be done remotely by a specialist, and the results of this checkup can be entered by the user with the help of the application's graphic interface. A lot of HAA have a built-in procedure of in situ-audiometry which automates the whole process so that the user can take this type of examination on their own. Software-based generators of tone signals and interface elements for reaction to exceeding the hearing perception threshold are used for that.

Quick change of settings, such as recruitment compensation, noise suppression, amplification regulation allows HAA users to choose the settings which are more suitable to understanding the speech in silence or with a background noise.

An additional feature of HAA is ability to choose and apply the sets of parameters most suitable for the current acoustic situation, for example, for silence, for talking in a noisy environment, etc.

Some HAA provide for using different computing formulas for calculation of target amplification on the basis of audiometry data. These formulas are intended to amplify the speech up to the maximum comfortable level of perception by the user.[2]

In particular, Pertalex HAA offers its users 3 well-known formulas:

• NAL-R;[11]
• Berger;[12]
• Pogo.[13]

When using microphone equipment of phone headsets, maximum available amplification is limited by appearance of acoustic feedback between the microphone and the speaker. Increase of possible level of signal amplification is ensured by stronger suppression of acoustic feedback.[14]

Acoustic feedback is the most widespread option of feedback appearing at the return leakage of sound from the speaker to the microphone. This can be caused by small distance between the microphone and the speaker, loose fit of an earpiece to the surface of acoustic meatus and so on.

The processing under frequency-dependent amplification of input signal for compensation of hearing impairment includes assistive functions, such as acoustic feedback suppression.[2]

In a number of HAA a scheme with subband signal decomposition is used for acoustic feedback suppression.[15]

Compared to traditional hearing aids, HAA has the following disadvantages:

1. more noticeable and not so comfortable to wear;
2. due to the fact that the microphone is not located in the ear, it does not use the functional advantages of the auricle and the natural acoustics of the outer ear.

At the same time, HAAs has a number of great advantages:

1. the large distance between the microphone and the speaker prevents the occurrence of acoustic feedback, which allows the use of large acoustic amplification and a simpler audio signal processing algorithm;
2. due to the rather large size of the device, it is possible to implement more convenient application control functions for people with poor motor skills;
3. using of various types of headphones is possible;
4. it is possible to achieve the highest sound pressure level and get high sound quality due to large speakers and a long battery life;
5. resistant to ingress of earwax and moisture;
6. thanks to the capacious battery, it is possible to use more complex audio signal processing algorithms and a higher sampling rate;
7. software flexibility;
8. developed software distribution system through appropriate services; the possibility of applying hearing correction algorithms to audio and video recordings, telephone calls, etc., and not only to acoustic signals;
9. the set up HAA in simple cases does not require special equipment and qualifications, the user can perform it independently;
10. HAA does not cause any psychological inconvenience, since the smartphone/tablet is not associated with the auditory pathologies in other people;
11. the user does not need to purchase and carry a separate device.

Of the 11 apps tested in the hearing assistant category, all 11 apps were still accessible at the end of the scoring period. The best app in the hearing assistant category was Petralex Hearing Aid (4.3/5). [16]

Distribution of HAA is limited by the infrastructure of specialized application stores, such as App Store, Google Play, etc. There is a fairly large number of applications that implements a hearing aid based on a smartphone/tablet. However, despite the general idea - hearing aid, they all differ in a number of criteria.

For benchmarking, the following applications were selected from the top 10 Google Play store applications (for Android devices) and the App Store (for iOS devices), selected on the "Hearing aid" request, and meet the following criteria:

• the release date of the latest version - no later than June 2012;
• the correspondence of the main functional declared by the developers in the descriptions solved by the application to the task - compensation of hearing impairments.

Comparative characteristics of hearing aid applications presented on the market are presented in the following table:[17][18]

Notes: *tested on iPhone 6+ (iOS 9.2), ** tested on Samsung Galaxy A3 (Android 7.0), *** on the date of testing (11 February 2019 (2019-02-11)), **** in the paid version, ***** - with the help of an additional application.

The principles of working HAA are similar to those of traditional hearing aids.

Most HAA provides two modes: settings mode (passing through the in situ-audiometry procedure) and the hearing correction mode.

Setup mode is an in situ-audiometry procedure that allows the user to independently measure hearing thresholds using tone audio signals that are generated with a gradually increasing amplitude over a period of some seconds. Audio signals are generated according to the following sequence of frequencies: 125, 250, 500, 1000, 2000, 4000, 6000 and 8000 Hz.

Some HAA do not have an integrated audiometry system. In this case, the adjustment of the sound amplification level (similar to the equalizer) is made by the user based on their own subjective feelings.[9][10][19][20]

Hearing correction mode represents an audio signal processing system. The audio signal processing system implements a broadband method for changing the correction of the spectral envelope of a signal using a filter with a finite impulse response whose frequency response is generated based on user-defined hearing thresholds during the in situ-audiometry procedure.

Figure 2 shows a general signal processing scheme in a hearing aid application.

Figure 2 - HAA general signal processing scheme

Audio input signal, acoustic feedback (echo) and background noise are captured by the microphone, summed and then divided into sub-band components with a lower sampling rate using an analysis filter bank (AFB).

Estimation of noise is based on the minima controlled recursive averaging of past values of noise.[21]

Cleared of noise subband signal is cleared of acoustic echo by adaptive filtering.[21]

The subband audio components that are cleared of echo and noise are amplified by multiplying by the corresponding coefficients, which are calculated based on the level of input and output audio signals, ambient noise energy, as well as hearing thresholds (hearing loss frequency characteristics) of the user.

The gain unit also performs the function of dynamic range compression, thereby compensating for the function of non-linear amplification of the human cochlea.

The processed broadband signal is synthesized using a synthesis filter bank (SFB).

AFB and SFB are can be realised as DFT-modulated filter banks, which is one of the most popular filter bank type used in modern hearing aids.[21]

The output signal can be multiplied by the total gain, which provides a comfortable sound level. This coefficient is adjusted by the user using an external controller directly in HAA interface.[9][5][6][7]

HAA, as a rule, provides high speed (i.e. minimum sound delay) and high sound quality due to a simple signal processing algorithm as well as due to the effective implementation of application functional blocks.[21]

Some HAA may not have an acoustic cancellation block.[22]

Based on the results of studies conducted on the effectiveness of using the different hearing aid applications, it can be concluded that for those respondents who have never used hearing aids before, there is a tendency of a positive effect on speech intelligibility. The positive impact on speech intelligibility when talking in a noisy environment was noted by those respondents who used hearing aids.[23]

HAA also can be useful not only as a hearing aid. For example, it can be useful in the process of studying, as students can place their smartphone/tablet closer to the teacher and hears him better.[8]