One Health: Improving access to hearing assessments leads to more data and better treatments

WEST LAFAYETTE, Ind. — Untreated hearing loss is associated with a slew of problems — such as cognitive decline, dementia and social isolation — and there’s no perfect fix. Research can help to better fit hearing aids and develop therapies to restore hearing, but researchers first need hearing loss profiles to include data linking specific damage hidden inside the ear, which isn’t part of an ordinary hearing test. To break the impasse, a team at Purdue University is developing computational methods for precise diagnoses that pinpoint the source of hearing loss in each individual, and accessible hearing tests to collect the required data from a more representative population, opening the door to innovative therapies for hearing loss.

“Purdue can meet the urgent need for precision audiology to ensure clinical trials and eventual treatments are targeted to the correct patients,” said Michael Heinz, associate head of research and professor of speech, language, and hearing sciences in the College of Health and Human Sciences, with a joint appointment in the Weldon School of Biomedical Engineering. Heinz is also the director of Purdue’s new Accessible Precision Audiology Research Center (APARC) in Indianapolis.

Heinz and a team of researchers have built a translational research approach combining foundational cross-species research, advanced data and AI techniques, and affordable open-source technology. Their work on cross-species studies of sensorineural hearing loss (which includes dysfunction in the cochlea or auditory nerve) is currently supported by a $3 million R01 grant from the National Institutes of Health.  In a partnership between Purdue and Creare LLC engineering research and development firm that is supported by a $3 million federal Small Business Innovation Research grant, the team is developing, adapting and validating a low-cost, open-source platform for cross-species hearing research. The approach builds on Creare’s open-source Tympan hearing system to assess hearing loss. Combined with a tablet computer and soundproof headphones, the Tympan can offer a fully mobile hearing test, vastly expanding access at a fraction of the cost in a world with only one audiologist for every million people.

In recognition of this promising approach, the university awarded the team a seed grant through a life and health science summit held in November 2023. The grant helps the team accelerate their work by expanding research facilities, creating a publicly available database of hearing data, and using AI and machine learning methods to chart the relationship between hearing assessments and internal damage.

“We want to create a marketplace where researchers can exchange data, models and tools, with the goal of understanding the etiology of hearing loss and best interventions,” said Ananth Grama, the Samuel Conte Distinguished Professor of Computer Science in the College of Science, and director of the Institute for Physical Artificial Intelligence. Computational tools developed for the problem — in areas of software infrastructure, signal processing, learning, causal inference and root cause analysis — contribute to Purdue’s One Health presidential initiative involving research at the union of humans, animals and plants.

“This work hits all of the benchmarks of a great research project for me: it stretches from the lab, augmenting learnings in the field, to helping ordinary people in the clinic; and there are far broader applications for the algorithms and technology they are developing,” said Kaethe Beck, Purdue assistant vice president for strategic impact in life and health sciences. The project, titled Data-Driven Cross-species Auditory Neuroscience for Accessible Mechanistic Precision Audiology, is one of four seed grant awards.

The research builds on Purdue’s top 10 audiology graduate program, two state-of-the-art audiology facilities — including APARC in Indianapolis and the Audiology Research Diagnostics Core in West Lafayette — and decades of basic laboratory auditory neuroscience research. With a goal of expanding hearing research to include underserved populations, Heinz is enthusiastic about the potential of APARC, which sits adjacent to a low-resource neighborhood of Indianapolis in the 16 Tech innovation district. Likewise, a hearing test using the Tympan could reach rural areas, where studies show that 50% of male farmers have a hearing handicap by age 50. The goal, Heinz said, is to “reach places where audiology isn’t readily accessible, improve diagnoses and treatments, and raise awareness about the importance of identifying hearing loss.

“Right here in Indiana, there’s a lot of farming communities where people have a higher prevalence of hearing loss and are also super busy and not next to an audiology clinic, so they tend not to be seen and not even know they have hearing loss,” Heinz said. “Or maybe they know, but they figure, ‘That’s fine, I still do OK at the quiet dinner table,’ because they don’t know about higher rates of other health issues that are associated with untreated hearing loss.”

Broadly speaking, four main subtypes of damage cause sensorineural hearing loss: damage to the inner- and outer-hair cells within the cochlea, degradation of a cochlear “battery,” and loss of synapses that transmit sensory information to the brain. Each will require a different treatment approach. And any one person might sustain a mix of the four subtypes. The result of a standard clinical hearing test, an audiogram, charts a patient’s hearing loss in terms of inability to detect specific frequencies at soft intensities, but doesn’t diagnose which parts of the cochlea are damaged and to what extent. And that does make a difference.

“Two patients can come into the clinic and have the same hearing loss, but with amplification one can understand speech and the other can’t,” Heinz said. “So, we know there’s major individual differences that aren’t being addressed.”

Diagnoses by subtype are not currently possible so drug companies looking to treat a subtype — for example loss of synapses in the cochlea — can’t even assemble a group of patients with hearing loss that shares an anatomical cause to take part in clinical trials. Accurate diagnosis alone would be a victory, Heinz said, in that it would both further encourage pharmaceutical research, and likely improve the method audiologists use to fit a hearing aid.

The team is looking at additional tests — measurements of middle-ear muscle reflexes, inner-ear acoustic emissions in response to sound, and neural activity — that could be used in concert with audiograms and cross-species data to establish patterns of cause and effect between damage and hearing assessments, and ultimately speech understanding in noise. With this approach, they will amass a large multimetric dataset perfectly suited for AI, Grama said.

“This becomes a machine learning problem where, if you see certain kinds of deficits or attributes in this set of hearing assessments, then we can predict the likely etiology,” Grama said. “And once we have a diagnosis, we can be a lot more precise in being able to derive a solution.”

In addition to Heinz and Grama, the research team also includes Edward Bartlett, associate dean and professor of biological sciences in the College of Science; Jennifer Simpson, interim department head and clinical professor of speech, language, and hearing sciences in the College of Health and Human Sciences; Joshua Alexander, associate professor of speech, language, and hearing sciences in the College of Health and Human Sciences; Maureen Shader, assistant professor of speech, language, and hearing sciences in the College of Health and Human Sciences; and Alexandra Hustedt-Mai, research audiologist in speech, language, and hearing sciences in the College of Health and Human Sciences.

About Purdue University
Purdue University is a public research university leading with excellence at scale. Ranked among top 10 public universities in the United States, Purdue discovers, disseminates and deploys knowledge with a quality and at a scale second to none. More than 107,000 students study at Purdue across multiple campuses, locations and modalities, including more than 58,000 at our main campus in West Lafayette and Indianapolis. Committed to affordability and accessibility, Purdue’s main campus has frozen tuition 13 years in a row. See how Purdue never stops in the persistent pursuit of the next giant leap — including its comprehensive urban expansion, the Mitch Daniels School of Business, Purdue Computes and the One Health initiative — at https://www.purdue.edu/president/strategic-initiatives.