In a number of studies, a test called AMTAS (Automated Method for Testing Auditory Sensitivity) has been proposed for automated recording of pure-tone audiograms which includes both the air-conduction and bone-conduction thresholds. In this method, the patient uses ‘Yes’ and ‘No’ buttons on a touch screen to indicate whether she/he is hearing or not hearing an audio signal. The signal level differs according to the responder's response to determine the level of hearing threshold of the patient. When generating a signal, sound masking is produced on the ear that is not tested. Circumaural transducers are used in this test to reduce the level of ambient noise. This enables the test to be done in a quiet room instead of a soundproofed room [278910].

In a study conducted by Margolis et al. [8], 30 participants (5 people with a normal hearing status and 25 individuals with hearing impairment) took part in the AMTAS test. The results showed that for air-conduction thresholds, the differences between the test values in the traditional method and the AMTAS were approximately similar to the differences in the values reported by two audiologists. However, for the bone-conduction thresholds, the differences between the recorded values with the AMTAS method and the manually recorded values were larger than the differences between the values reported by the two audiologists. Two reasons mentioned for these differences were incorrect reference-equivalent threshold force levels for bone-conduction through the forehead bone and a differential effect of middle ear diseases on the forehead and mastoid bone-conduction thresholds. Some studies showed that forehead bone-conduction thresholds are less affected by middle-ear diseases than mastoid thresholds for patients with conductive hearing loss. In another study conducted by Eikelboom et al. [2], the AMTAS test was performed in a quiet room for 44 participants with different levels of hearing impairment. The results indicated that the audiometry changes with the air-conduction and bone-conduction methods were within an acceptable level for the automated and manual methods. Although the AMTAS thresholds were higher compared to the manual methods, no significant difference was reported.

In another study, a software program called the Home Hearing Test (HHT) was developed to record the air-conduction thresholds at home. The purpose of this study was to compare the results of the tests performed by the patients with the results of the automated audiometry in a clinical setting. The difference between the HHT and manual thresholds was slightly higher than the recorded thresholds by two audiologists and the measured difference between the AMTAS results and the manual method in a clinical setting. Some of the reasons for this difference were a long time interval between the HHT test at home and the manual test in a clinic (53 days), the probability of environmental noise affecting the thresholds of the HHT and the differences among the participants in terms of the severity of hearing impairment [10].

The software solution is not limited to AMTAS, and different software has been developed for automated audiometry. For example, in Poland, web-based audiometer software was developed. Three tests were performed to evaluate the software: a manual test, an audiometry test under the supervision of an audiologist in a sound insulation room, and a test which could be done by the patient at home. There was no limitation for the type of earphones used to do the test at home. The results revealed that a web-based audiometer can be used in screening tests. Although performing audiometry tests requires basic knowledge in the context of hearing thresholds or frequencies, a user friendly interface can be used when an audiologist is not available [11].

Another study was conducted in the United States in which a web-based distributed pure-tone hearing assessment system was developed. It had a three-layer architecture, which increased the scalability of the system to be integrated with other audiometry services. Moreover, the audiometry data were stored in a standardized database, which could be integrated with Electronic Medical Records. In terms of clinical effectiveness, the results showed that the web-based software worked similar to the traditional method at all frequencies. Moreover, the bandwidth required for the system was less than 1 MB/s [12].

In a number of studies, hardware solutions have been proposed to be used in automated audiometry. According to the literature, the KUDUwave portable audiometer [131415] and an earphone designed to remove environmental noise [16] were among the hardware solutions proposed for performing automated audiometry. The KUDUwave is a portable audiometer that makes real-time audiogram recording possible. This audiometer controls the noise attenuation by using earphones and circumaural ear-cups and provides an opportunity for performing hearing tests down to zero dB with an environmental noise up to 59 dB SPL. Moreover, the environmental noise levels are continuously monitored, and if the environmental noise level exceeds the limit, the audiometry test will be stopped. Therefore, this test can be performed outside a sound insulation room [13]. The KUDUwave uses sound masking when it is needed. If the difference between the air conduction thresholds in the test and non-test ear is 75 dB or more when the frequencies are smaller than or equal to 1,000 Hz, or if the difference is 50 dB or more when the frequencies are larger than 1,000 Hz, a masking level of 30 dB is used above the non-test ear. For bone conduction thresholds, a continuous masking level of 20 dB is used above the non-test ear [1314].

In a study conducted by Swanepoel and Biagio [13], the performance of the KUDUwave was evaluated for 30 individuals aged 19 to 77 years. The results indicated that the air-conduction thresholds had a difference of around 5 dB from the values recorded by the traditional method for 90% of the participants. The bone-conduction thresholds had a difference of around 10 dB from the values recorded by the traditional method for 92% of the participants. However, all the values reported by the KUDUwave were clinically acceptable. A small difference was seen in the test-retest. The results of the test-retest were approximately the same in 91% of cases when the KUDUwave audiometer was used and the difference range was 15 dB. For the traditional audiometer, the results were approximately the same in 92% of cases with a difference range of 10 dB. The reason for the small difference was unknown. However, it might be due to the bone vibrator attachment to the circumaural earphones that produces more static forces in the displacements [13].

In another study in Australia, the hearing status of people was evaluated by using the KUDUwave in the traditional and automated tests. The traditional test was done in a soundproof room, and the automated test was conducted in a room without sound insulation. The bone vibrator was placed on the mastoid bone in the traditional test, while it was on the forehead bone in the automated test. These factors, along with environmental noise, could affect the results. However, the results showed that the difference in the hearing thresholds was low. The results suggested that 86.5% of the recorded thresholds were in a 10-dB range. Although the differences in the thresholds became statistically significant, further research is needed to identify whether these differences are clinically meaningful [14].

The KUDUwave audiometer was also examined in South Africa. The air-conduction hearing thresholds of 30 participants with normal hearing status and 8 participants with hearing impairment were recorded using both automated and traditional methods. The results showed that the automated audiometry is a stable, accurate, and time efficient method to evaluate the hearing status of adults with normal hearing or hearing impairment. The combination of the automated audiometry with an asynchronous telehealth model, especially for poor areas with little access to hearing specialists, would be beneficial to improve health care services [15]. In a study conducted by Meinke et al. [16], a mobile wireless automated hearing-test system (WAHTS) was designed to reduce the environmental noise and to be used to record the hearing thresholds in a non-sound proof environment. The system performance was evaluated by examining the air-conduction thresholds of 20 workers in six locations, and the results were compared with the results of another test conducted by using computer-controlled audiometry in a mobile trailer sound booth. Overall, the difference between the thresholds obtained by WHATS and the thresholds obtained in the mobile trailer sound booth was within 5 dB [16].

In a study conducted by Whitton et al. [5], a tablet-based application was developed in the United States. The test algorithm followed the same rules of the clinical test. The audio tones were provided for a time interval of 3 to 7 seconds, and the participants' responses were considered to be a correct answer even up to 2.5 seconds later. The test was conducted in a home and clinical environment. The difference between the mean values recorded at home and at the clinic was small. The thresholds recorded at home had increased for very low frequencies (≤250 Hz). This increase could be due to an increase in the ambient noise at low frequencies in a home environment. This study showed that it is possible to monitor hearing impairments outside a clinical environment.

In another study, a version of the hearScreen smartphone app was used to record the hearing thresholds. The exclusion criteria was a unilateral hearing loss of more than 40 dB HL to avoid inter-aural effects because contralateral masking was not considered in the prototype of the smartphone app. Conventional thresholds exceeding 15 dB HL corresponded to smartphone thresholds within ≤10 dB in 80.6% of the cases. This study suggested that air-conduction audiometry can be performed accurately by a smartphone-based audiometer in a soundproof room or outside a sound insulation room and in the healthcare clinics of poor areas [3].

The hearTest application was another app developed for Android smartphone-based audiometry. This app was used with a supra-aural earphone. However, no significant difference was seen between the app results and the traditional thresholds except for a 4 kHz frequency. In total, 70.6% of the thresholds calculated by the app and the traditional method had less than a 5 dB difference. Moreover, the duration of the test was not significantly different for the two methods [17].

uHear audiometer was another application developed by Khoza-Shangase and Kassner [18] in South Africa, and its accuracy was compared with the traditional approach. uHear is an automated screening test which is downloadable for iPod and iPhone. The participants of the study were 86 primary school students. The differences between the results obtained from uHear and the traditional method were significant at all frequencies. In this test, non-calibrated insert earphones were used for automated testing. The advantage of insert earphones compared to supra-aural earphones is reducing environmental noise with greater accuracy. However, this advantage can only be obtained if the earphones are correctly and recently calibrated. Because the study was conducted in a school environment, environmental noise could affect the outcomes. That might be the reason why uHear was not as accurate as the traditional audiometry in determining the hearing thresholds for primary school children. This study showed that caution should be exercised in using uHear and that further research evidence is needed to use it at a general level.

Another iOS-based application was developed to be used on the iPhone, iPod touch and iPad. The hearing thresholds were recorded with an automated method in a quiet room as well as in a sound insulation room. Sound masking was applied automatically when the difference between the threshold values of the two ears was greater than or equal to 35 dB. The sound masking included a narrowband noise centered at the frequency that was tested. The results showed that on average, 96% of the thresholds recorded in the sound insulation room with the automated test had a difference in the range of 10 dB compared to the thresholds recorded in the sound insulation room with the traditional method. In a quiet room, this amount reached 94%. The results indicated that the obtained thresholds were close to traditional audiometry results, and the iOS-based tools provided a platform for conducting automated audiometry with no need for additional equipment [19].

Having reviewed the literature, it was revealed that there are two important factors regarding the choice of earphones or their design when using them for audiometry tests. The first factor is related to reducing environmental noise, and the second factor is related to the occlusion effect of the earphone. It is notable that the background noise in the test room is called environmental noise. Because the automated audiometry test might be performed in an environment outside a sound insulation room, the environmental noise levels must be minimized to be able to record the hearing thresholds precisely. The ambient noise level should be much lower than the test signal level, so that the listener can distinguish the test signal from the environmental noise [16]. The occlusion effect causes a change in the values recorded for the bone-conduction thresholds (usually an incremental change) due to the obstruction of the ear canal. The skull vibration is transmitted to the walls of the external ear canal and tympanic membrane. When the ear canal is not clogged, the ear canal acts as a high-pass filter, and low-frequency sounds are removed. When the canal is clogged, the low-frequency energy falls to the trap and transmitted into the inner ear [20].

The standard procedure for measuring the bone-conduction thresholds is to perform the test in a condition in which the ear canals are not clogged. On the other hand, earphones are used to record the air-conduction thresholds. If the earphone does not produce the occlusion effect, the earphones can be kept on the ears during the test. This feature enables recording the air-conduction and bone-conduction thresholds with no need to switch earphones and causes no interruptions during the test [7]. In general, noise reduction techniques improve the value for hearing thresholds for low frequencies by using two active and passive modes. The passive technique attempts to prevent the environmental noise from entering the ear canal. In the active noise elimination technique, a microphone is used to measure the amount of noise in the environment, which is neutralized by using environmental noise opposing signals [5].

The available earphones can be divided into three main categories: circumaural, supra-aural, and insert earphones. Figure 2 shows how each earphone couples to the external ear. The KUDUwave audiometer is powered by circumaural ear cups with insert earphones to control the environmental noise. Moreover, there are microphones in the inner and outer parts of the earphones to monitor the ambient noise, and if the level of ambient noise exceeds the limit, the audiometry test is stopped. These features help to perform the evaluation with a better quality in an environment outside of a soundproof room [13].

In a study conducted by Meinke et al. [16], a headset was designed to reduce the environmental noise and to be used to accurately record the hearing thresholds outside an insulating room and in an industrial environment. The headset was relatively large and heavy to do passive attenuation. It had multiple layers to eliminate environmental noise by a passive technique. The ear cup with thick polyurethane foam, a speaker, a plastic face plate, a thin protective fabric and an ear seal from hearing protectors were lined to eliminate noises. The test-retest reliability results were equal or better than the results from the supra-aural, insert and circumaural earphones.

In another study, the occlusion effects created by earphones were highlighted. One of the important points in performing an automated audiometry test is to prevent the occlusion effects created by earphones. Supra-aural earphones produce a significant occlusion effect at frequencies of 1 kHz or lower. Insert earphones also produce a significant occlusion effect unless completely inserted into the ear canal and into its bony part. Putting the earphones in this way is not possible in routine tests due to the lack of comfort and safety; however, if the earphone is put around the ear and a large volume of air is placed inside, it can eliminate the occlusion effect. The results of this study showed that the occlusion effect produced by circular earphones for frequencies above 500 Hz is so insignificant, and the thresholds recorded in the bone-conduction test with this type of earphone are similar to the thresholds recorded without earphones. Therefore, according to the results, the circumaural, insert and supraaural earphones had a better performance in not generating the occlusion effect, respectively [7].

One of the important technical factors in an automated audiometry test is the location of placing the bone vibrator. The vibrator is used to generate signals for determining the bone-conduction thresholds. The two common methods for producing these signals are the use of the mastoid bone and the forehead bone. The mastoid bone is used mostly in traditional audiometry techniques [7]. In the case of using the mastoid bone, the vibrator needs to be placed on the left and right ears and between the ears. To increase the efficiency, it is better not to move the oscillation tool from one ear to another ear. The forehead bone is preferred in automated testing, since there is no need to change the vibrator placement during the test [8].

The KUDUwave audiometer uses a bone vibrator that is placed on the forehead bone. In this audiometer, the insert earphones are used during the testing of the bone-conduction thresholds. As previously mentioned, the earphones should be deeply inserted into the bone part of the ear canal to avoid the occlusion effect of insert earphones. According to the results, the correlation between the bone-conduction thresholds in the test-retest is slightly less than the measured values in the traditional audiometry tests, although they were within acceptable limits. The reason for this small difference was unknown. It might be partly due to the bone vibrator which was attached to the circumaural earphone and generated more static force in the movements [13].

In the case of using the forehead bone, the earphones are used on both ears. When testing each ear, sound masking is produced on the other ear. Therefore, it is important that the earphones create a small amount of obstruction to have no effect on the bone-conduction thresholds. The difference between the two traditional and automated methods in the bone-conduction thresholds is greater than that of the air-conduction mode. This difference can be due to the difference in the position of the vibrator on the mastoid bone and the forehead bone or due to middle ear diseases, which affect the sensitivity of the bone-conduction thresholds in both the mastoid and forehead methods. The forehead method estimates the cochlea sensitivity with a higher accuracy [8].