Validation of a Smartphone-Based Hearing-in-Noise Test (HearMe)

Age-related hearing loss, or presbycusis, is highly prevalent among the elderly. It can have a major impact on the quality of life due to progressive communication difficulties that may lead to psychosocial dysfunction and in extreme cases to social isolation and depression. Hearing aids and cochlear implants have evolved rapidly over the past decades and may strongly improve quality of life in elderly with hearing loss. Nevertheless, it is estimated that in the US and Europe only one out of five to six people with substantial hearing loss is actually using hearing aids. In fact, nearly 50% of people with hearing loss never underwent a hearing test at all. This may be due to denial of illness associated with aging or to disbelieve in bene t from hearing aids or cochlear implants. Hearing screening could increase awareness of hearing loss among elderly and identify those who might bene t from amplification. However, classical audiometry (i.e. pure- tone thresholds and speech-recognition in quiet) is not ideal for screening purposes as it is time consuming and it requires expensive equipment including a soundproof booth and a calibrated audiometer.

Over the past decades, several Speech-In-Noise tests were developed to get a better assessment of a person's hearing ability in real-life situations. These tests generally measure the speech reception threshold (SRT) in dB signal to noise ratio (SNR). SRT is determined as the difference between the level of presented speech and background noise at which the tested per- son can correctly reproduce 50% of words or sentences. It is now generally recognized that the SRT is more representative of a patient's hearing ability in real-life situations than pure- tone audiometry or speech recognition in quiet. The Speech- In-Noise test, based on spoken sentences, is still being used in clinical practice. However, its general use for a broad clinical population has been disputed because not every person is able to understand and repeat complete sentences in noise, regardless of his hearing loss. This test is there- fore considered as an assessment of the entire auditory system, including memory and certain linguistic aspects, rather than of hearing loss alone.

In 2013, the Digits-In-Noise (DIN) test was developed, overcoming several shortcomings of the previous telephone test. It uses, for example, wideband signals instead of the limited telephone bandwidth and concatenating digits, spoken by a male voice. The DIN test requires listeners to repeat three spoken numbers (a so-called digit triplet) presented through a headphone, while a continuous noise is presented synchronously to the same ear. The response is then scored correct or incorrect automatically by the computer. Depending on the test setting, the response can either be imputed by the listener him- self or by an administrator. By using simple digits in a closed set paradigm, the contribution of top-down processing and thus the influence of cognitive status is thought to be minimized. For this test, no learning effect was detected, low measurement errors were reported (0.7 dB SNR for normal-hearing listeners) and high validity was claimed by comparing measured digit- triplet SRTs and sentence SRTs. Although SRT-in-noise reflects different aspects of hearing acuity than pure-tone thresholds, both measures are also highly correlated. Previously reported correlation coefficients vary from 0.77 to 0.86. These studies included populations with a wide range of hearing losses, varying from severe to no hearing loss at all.

Based on this relationship, the DIN test could potentially be used as a screening instrument for hearing loss. However, these two studies did not report specifically on a population of elderly subjects, and this is one of the main target groups for hearing screening. It is important to validate the assumed relationship between SRT-in-noise and hearing loss for this particular population with a smartphone, as it may be influenced by general aging, for example, a decline in cognitive skills.

The purpose of this project is to validate a quick, easy-to-use and administer smartphone hearing-in-noise test. The app (called HearMe) can potentially be used to easily and quickly collect hearing-in-noise and speech-in-noise measurements. The smartphone app developed is a hearing-in-noise test that presents the subject with a series of stimuli consisting of a spoken three-digit sequence presented at a varying hearing-to-noise ratio. For each stimulus presentation, the user tap the three-digit sequence. The duration of the app is less than 3 minutes. For this project we will test at least 50 participants with hearing loss and 50 control subjects between the ages of 18-80. The participants will be invited to take the app. The approach for this pilot study is to characterize hearing-in-noise thresholds (also referred to as a speech-reception threshold) as measured by the app in both participant groups, and relate it to the phenotype of each group as a preliminary evaluation of the app as well as a preliminary validation against their routinely collected measurements of hearing function (pure-tone audiometry thresholds).

With this study, the investigators aim to evaluate the developed smartphone HINT/DIN test for its ability to screen the elderly for hearing loss. The investigators hope to examine the relationship between pure-tone thresholds and SRT-in-noise as measured by the HearMe smartphone application.