Hearing Loss, Hearing Loss, Sensorineural
Conditions
Brief summary
The goal of this clinical trial will be to evaluate the efficacy of artificial intelligence-integrated hearing aids in individuals with hearing loss. The main questions to answer are: 1. How effective is an artificial intelligence integrated hearing aid in improving speech perception in noise. 2. How does an artificial intelligence integrated hearing aid compare to currently available commercial hearing aids.
Detailed description
Hearing loss is a leading cause of disability globally. Undertreated hearing loss has been associated with several deleterious sequelae, including social isolation, decreased quality of life, and cognitive decline. For a large portion of individuals with hearing loss, hearing aids remain the intervention of choice for improving communication, decreasing social isolation, and enhancing quality of life. Yet, less than 20% of hearing aid candidates choose to use them2. This is partly due to a perceived lack of benefit from hearing aids. Individuals who use hearing aids continue to report difficulty with speech perception in noisy environments with multiple speakers, such as parties, restaurants, and other complex auditory settings. The investigators have developed a neural network that runs on-device to achieve real-time target speech hearing in realistic multi-talker environments. Using our neural network, the wearer can focus on speech from a target speaker by learning their unique speech cues while ignoring all interfering speech and noise in complex acoustic environments. The wearer is additionally able to use distance to separate sounds from the environment from target sounds to be amplified. Finally the wearer is also able to specificy categories of sound that will be amplified in environment. This study will prospectively study a hearing aid implementing these algorithms on individuals with hearing loss. Participants will be randomized to utilizing one of two commercial hearing aids vs the experimental device. Participants will participate in listening tests in which they must identify the words spoken by a specific voice in multi-talker babble noise. The investigators will measure performance with the primary and secondary outcome measures listed elsewhere.
Interventions
DEVICEArtificial Intelligence Integrated Hearing Aid Use
Participants in this arm will use the study-designed artificial intelligence-integrated hearing aid in the listening tasks. Participants will also perform the listening tasks without hearing aids to act as controls.
DEVICEPhonak Infineo Hearing Aid
Participants in this arm will use a fitted commercial hearing aid in the listening tasks. Participants will also perform the listening tasks without hearing aids to act as controls.
DEVICEOticon Real Hearing Aid
Participants in this arm will use a fitted commercial hearing aid in the listening tasks. Participants will also perform the listening tasks without hearing aids to act as controls.
Sponsors
University of Washington
Study design
Allocation
RANDOMIZED
Intervention model
PARALLEL
Primary purpose
TREATMENT
Masking
NONE
Masking description
No masking will be performed
Intervention model description
Participants will be randomly assigned to groups for each hearing aid device being tested (e.g. commercial model 1, 2, vs study device)
Eligibility
Sex/Gender
ALL
Age
18 Years to No maximum
Healthy volunteers
No
Inclusion criteria
* Minimum Age of 18 yo * Bilateral sensorineural hearing loss amenable to hearing aids * Able to give informed consent * Native English speaker * Able to wear a standard pair of hearing aids
Exclusion criteria
* Unable to participate in informed consent * Use of a cochlear implant * Hearing loss not amenable to hearing aid use * Conductive hearing loss
Design outcomes
Primary
| Measure | Time frame | Description |
|---|---|---|
| Speech Reception Threshold | From enrollment in the study until completion of a one day listening test experiment. | A measurement of difference in average speech reception threshold across all groups |
| Awareness of Environmental Marker | From enrollment in the study until completion of a one day listening test experiment. | During certain listening tasks, an environmental sound of concern (e.g. a siren, or alarm) will be played while the participant is tasked with identifying speech. Awareness of these sounds will be tested by reaction time to the sound and identification of the sound. |
Secondary
| Measure | Time frame | Description |
|---|---|---|
| Preference Scores | From enrollment in the study until completion of a one day listening test experiment. | Mean opinion scores about which device is preferred |
| Sound quality | 5 minutes | Mean opinion scores about subjective sound quality of respective devices |
Contacts
Primary ContactArun Raghavan, MD
Outcome results
None listed