Artificial Intelligence and Machine Learning: Pushing New Boundaries in Hearing Technology : The Hearing Journal

Journal Logo

Cover Story

Artificial Intelligence and Machine Learning: Pushing New Boundaries in Hearing Technology

Wolfgang, Kelly

Author Information
The Hearing Journal 72(3):p 26,27,30, March 2019. | DOI: 10.1097/01.HJ.0000554346.30951.8d
  • Free

No human can do it, but the machine can,” said Fan-Gang Zeng, PhD, a professor of otolaryngology at the University of California, Irvine, School of Medicine, and the director of the Center for Hearing Research. Zeng, referring to artificial intelligence (AI) capabilities in hearing aids, may sound futuristic in his assessment, but as companies begin to integrate an array of new technologies, including machine learning, 3D printing, real-time language translation, and fall detection, to hearing aids, it is evident that the future of hearing care is now.

hearing aids, AI, machine learning, data


“Looking back at the last nearly 20 years of history in hearing care, you will find that technology and technological breakthrough have always been a part of the industry,” noted Thomas Lang, the senior vice president for Phonak. “But now, artificial intelligence has reached a status where its solutions have become relevant for hearing care through direct connectivity, the ability to read and process large amounts of data, and the capacity to make intelligent proposals for users,” he added. Such technologies have not only revolutionized the hearing capabilities of hearing aids, but also provided a unique opportunity to improve users’ lifestyles. One landmark breakthrough in hearing aid AI is the ability for devices to log users’ preferences in a variety of listening environments, from daily work spaces to concerts, movie theaters, restaurants, public transportation, quiet home environments, and more.

Now the hearing aid will monitor adjustments users make in each space, log those preferences, use GPS to detect when the user returns to those places, and either automatically adjust to preferred settings or send an app-based push notification to ask the user if she or he would like the device to adjust. “When it comes to analyzing the sound environment and recognizing what you like and don't like, AI does a great job of data logging and helping individuals,” Zeng said. “It's something current hearing aids cannot do and an area where the old statistic model of pattern recognition failed miserably,” he said. “AI provides a realistic possibility for hearing aids to improve speech recognition in noise, which has been the number one complaint of patients but is still challenging to current noise-cancellation technologies,” Zeng added.

Dave Fabry, PhD, the chief innovation officer for Starkey, added that machine learning technology ensures that hearing aid quality and performance in both quiet and noisy environments provides an exemplary experience using personalization and optimization. “[Automatic preference adjustment] will provide satisfaction not only in terms of understanding speech but also with spatial sounds around you,” he said.

Although similar technology is available in other consumer electronic devices, Lang noted that the introduction of AI and ML in hearing aids brings the vision for everyone to enjoy the delight of better hearing without limitations to fruition. “Now hearing aids are compatible with the devices that already use AI and can deliver quality listening depending on the use case even when listening to music through smartphones or watching television,” he said. “Now, we are able to combine connectivity, rechargeability, and superior sound quality to bring everything together in one product,” Lang added.

AI has also introduced the ability for hearing aids to detect human speech, perform real-time language translation, monitor physical and mental health, and detect falls. Using machine learning, hearing aids have the ability to access data and ultimately predict them, making automatic speech detection and real-time, amplified translation of foreign languages directly in the ear possible. Incorporating motion sensors also enables users to track steps without the use of a separate device. The same technology allows hearing aids to detect falls and even send notifications and GPS location to pre-determined people such as family members and/or care providers that the user has fallen.


Further technological developments including 3D printing have allowed manufacturers to incorporate AI into hearing aids that are more comfortable and sleeker than past models. “When building a product that goes into the ear, we rely on hearing care professionals to start the process by taking impressions of each patient's ear canals,” Lang said. “We then use a modeling process where the device is built on a computer and a printed shell is produced on a 3D printer. We push the technological advancements further by using medical-grade titanium to build the shell, rather than plastic and acrylic materials.”

The 3D printers used to produce hearing aids utilize titanium powder and a laser beam to build the shell layer by layer, resulting in a hearing aid that is stronger, tougher, and thinner than the plastic devices, with deeper penetration into the ear and a more discreet appearance.


Introduction of such technologies through AI and ML has the ability to improve the everyday lives of users and reduce common comorbidities of hearing loss, such as social isolation, dementia, and cardiovascular disease. “Now, AI hearing aids can log how often a user is wearing a hearing aid and whether he or she is engaged in conversation with other human beings,” Fabry said. “Research has shown that particularly for the aging population, if those with hearing loss are not engaged, they become withdrawn and depressed, which can accelerate the path of cognitive decline.”

Lang noted that when people with hearing loss start having issues with understanding, they often begin to skip family gatherings or events with friends. “Hearing loss has a known correlation with dementia, Alzheimer's, risk of fall, and hospitalization,” he said. “Often, there are many more things happening when people develop hearing loss. By focusing more attention on these topics, we aim to provide more than just amplification to hearing aid users.”

AI capabilities also provide an opportunity to focus on the physical activity of those with hearing loss, combatting known comorbidities such as sedentary lifestyles that often accompany aging and can lead to obesity and cardiovascular disease. “By incorporating technology and developing hearing aids with a purpose, such as step tracking, we can target those with an elevated risk of diabetes and stroke, for example,” Fabry said. Hearing aids with embedded step tracking allow users to exercise without additional devices and provide notifications to users to keep moving throughout the day and reach goals, such as the recommended 10,000 steps per day, he added.


For hearing care providers who work directly with the technology and the end-user, AI and ML provide the opportunity to introduce new features to their patients and practices. Though the emerging technology allows users to be more independent than ever, “nothing can replace the smiling face of a provider,” Zeng said. “Hearing aids are mainly used to communicate with people, not just for the added benefits AI can provide,” he added. “Technology will never replace the patient care aspect an audiologist or hearing instrument specialist can provide, and we are the first interface with patients.”

AI does, however, provide opportunities to interact with patients in new ways and provide facets for overall lifestyle improvements.

“Audiologists are not able to meet with patients at all times, and patients don't always want to go into the office for minor adjustments. One common challenge for auditory care is that initial fitting and adjustment of hearing aids are typically performed in a clean room,” Fabry said. “As soon as patients leave and go into the real world, they are faced with challenges such as hearing in their favorite restaurants or listening to their grandchildren,” he said. But AI advances allow users to initiate requests for adjustment from afar, where hearing professionals can make modifications to hearing aids using their clinical judgement over the cloud, he said. “Of course, these advancements don't replace the engagement between audiologists and their patients, but the ability to augment minor adjustments provides the innovation needed to achieve outcomes with patients better than ever before.”

Telehealth through AI also allows for remote fittings, Lang said. “Initial fittings are done in a practice and may require additional follow-up appointments,” he said. “Now, we have introduced the possibility for a hearing care professional to do follow-up fittings remotely so the consumer doesn't have to drive to the practice.” For practitioners, the process is straightforward, Lang noted. Providers need a simple fitting software and patients need a compatible cell phone; that technology allows hearing care professionals to provide a simplified onboarding experience of making appointments, performing remote fittings and support sessions, and interacting with patients as if they had been sitting in the clinic, Lang said.

“No one knows what is best for the patient more than the caregiver,” said Donald J. Schum, PhD, the vice president of audiology for Oticon. “We see these developments as a way for hearing care professionals to gather more personalized information about users, allowing for better crafting of the hearing health solution,” he said.

Clinical audiologist Jennifer Torres, AuD, CCC-A, finds this emerging AI technology exciting as it pushes audiologists to be better; however, it is important for hearing care providers to be aware of how the technology will fit into patients’ lives in a practical way. “We have found that the current population of patients with hearing loss in the 80-90 age range don't have the technology to support AI hearing aids, such as smartphones,” she said. “As hearing care professionals, we are the interface between the user and hearing aid smart devices. It can add frustration to the patient if she or he is not able to use hearing devices without the use of a smartphone. Even if the hearing aid is user-friendly, patients still need to understand apps to stay connected, and that can challenge management on the patient end.”

Torres added that the constant connectivity can be a great feature, but may pose a challenge to facets of practice management such as scheduling, billing, and insurance coverages. “Most audiologists still work the traditional 8-5 schedule, and may find it difficult to find some level of administrative time to deal with unscheduled remote requests or other tasks that make a huge positive difference for patients but aren't billable,” she said.


Though AI is revolutionizing the listening experience for hearing aid users, its greatest potential may be the ability to make hearing aids desirable, thereby removing the stigma of use and reaching more patients with hearing loss than ever before. “Anything that makes patients’ lives a little easier is going to improve the acceptance of that device, and ultimately, overall patient care,” Torres said.

“AI technology will transform hearing aids from a device used by the hearing-impaired to something everybody can use that will help their daily lives,” Zeng added.

Lang predicted that the introduction of AI and ML capabilities such as language translation and motion detection will expand the hearing aid market because more people will want to benefit from such innovative solutions. “With the ability to directly stream music and movies to the ear, it's now becoming cool to wear a hearing aid,” he said. “It's a big shift because traditionally, hearing aids are not something patients want but something they need. As AI evolves to cover more solutions and benefits, the stigma will disappear.”

“Increasingly, we are moving away from one-size-fit-all predicted solutions and toward solutions that are patient-centered, taking into account individual, social, psychological, and physical capabilities and needs to create a truly human-driven hearing health care model,” Schum said.

Fabry added that connected hearing aids are now multipurpose, multifunction devices that provide exceptional auditory benefits and have the ability to allow hearing aid users to become connected to the world. “It's fundamental that we never forget that we are not just working on technology that goes into hearing aids, but also working on providing better hearing and, in doing so, enabling people to live better in any stage of their lives.”

Thoughts on something you read here? Write to us at [email protected]

Copyright © 2019 Wolters Kluwer Health, Inc. All rights reserved.