Skip to main content
2024 BMES Annual Meeting Main banner
Icon Legend
This session is not in your schedule.
This session is in your schedule. Click again to remove it.

High School Student with a Traditionally Underrepresented Background

HS Poster

Poster 53 - Development and Clinical Evaluation of An Online Tinnitus Suppression Platform for Customized Sound Therapy

Saturday, October 26, 2024
10:00 AM - 12:00 PM EST
Location: Exhibit Hall E, F & G

Introduction: Tinnitus is the sensation of hearing sound in the absence of an external auditory stimulus. Despite being a relatively obscure auditory disorder, data shows that around 20 million people suffer from chronic tinnitus, with 2 million reporting "extreme cases" that severely impact their quality of life. While there is no miracle one-time cure, tinnitus can be suppressed through sound therapy. Ambient sounds such as bird noises, white noise, and ocean waves have been proven to mask sounds caused by tinnitus by covering the user's perceived frequency with a second audio stimulus: generating an umbrella effect. The selection of sounds is crucial for effectively suppressing tinnitus. A major challenge faced by all current tinnitus treatment models is the lack of accessibility to the general public. High costs and a lack of awareness about effective treatments have consistently prevented many patients from seeking the help they urgently require. This study aims to develop an online sound therapy platform with an automated self-testing capability through which patients can get matched to a recommended sound file that maximizes relief based on their frequency band preferences. Such a platform enhances the accessibility of real-time targeted tinnitus therapy, addressing a significant gap in availability and convenience.

Materials and

Methods: Our study revolved around implementing our customized sound therapy (CST) algorithm in an initial clinical trial. Materials used included a customizable cloud-based survey platform (SurveyPro), a data organization hub (custom-coded Google spreadsheet), 12 band-specific noises, 12 random tone sequences, and 50 files of audio sound collection. The first stage of the CST algorithm is a band-specific noise test to narrow the patient's tinnitus within 12 frequency ranges (362 Hz - 7132 Hz). Patients are then presented a random noise (oscillating sound 25% above and below the median value) test in the same 12 ranges for data collection. A patient would rate the effectiveness of tinnitus suppression for each band-specific sound with a score from 0 to 10, with 10 being the most effective. With the patient's rating results, we created our CST algorithm, which uses the Fast Fourier Transform (FFT) to match the specific patient with the best audio file out of the 50 sound files with various spectral profiles. The matching algorithm was developed by comparing the user's 12-band ratings with the spectral profiles of the 50 audio files through a minimum squared error (MSE) test.

Seventeen tinnitus patients (5 more in our pilot study) were recruited and tested through Shandong University Qilu Hospital's otolaryngology department. The ages of the patients range from 18 to 60+, and the lengths of tinnitus range from less than one year to 20+ years. Patients were admitted to the actual audio test in a clinical soundbooth under the direct supervision of a clinician.

Results, Conclusions, and Discussions: Our results included a multi-part analysis of collected patient data: comparing band-specific noise test performance against random tone sequence and measuring masking effectiveness of recommended audio files by comparing rating with spectral profile. Figure 1 presents the rating scores of individual patients for the custom-matched sounds. Of the 17 patient profiles we tested, 11 subjects reported a rating score of 7 or higher in response to the audio files our algorithm sourced for them (out of our bank of 50). 6 subjects out of that group of 11 gave their custom-matched audio a rating score of 10: completely masking their tinnitus. Although our group received promising results, there are three instances of perplexing user responses. Subjects 7, 11, and 13 all rated their matched audio lower than 7 despite having relatively strong matches between the spectral profile curves—of the 50 audio files—and the user ratings curves. A few likely causes are listed below.

Limited collection of sound files, ideal to increase beyond 50
Small sample size
The first round of 12 band-specific noise tests each covered too broad a range, not specific enough to narrow down each patient’s tinnitus frequency

See Figure 1

Conclusions

In developing an online tinnitus evaluation platform for customized sound therapy, we explored measuring tinnitus suppression using band-specific noise and multitone audios, evaluated the results that hold the potential for advancing effective tinnitus therapies, created selected audio spectral profiles for tinnitus sound therapy, and gained insights on what other factors may be conducted for more accurate results.

We generated narrow-band masking sounds and random tone sequences within specified frequency ranges. Using a multistage survey, patients rated audio frequencies from 362 Hz to 7132 Hz on a scale from 0 to 10 for tinnitus masking. A machine learning algorithm then selected the most matched sounds. Initial tests showed that 11 out of 17 subjects rated a matched audio with a score of 7 or higher for tinnitus suppression.

Our team aims to create a database system to recommend tailored audio treatments and inspire further exploration in this field.

Acknowledgements (Optional): We sincerely thank our research mentor, Dr. Kaibao Nie, for his valuable insights, support, and directions that helped move our developments and research regarding tinnitus suppression forward. We also sincerely thank the doctors and clinicians from Qilu Hospital of Shandong University, China, who were instrumental in conducting our study. We also wish to thank all patients who participated in our survey and provided invaluable data for our study.

References (Optional): Ganesan, P., Rodrigo, H., Schmiedge, J., Reiter, R., Swapna, S., & Manchaiah, V. (2021). Combined Amplification and Sound Therapy for Individuals With Tinnitus and Coexisting Hearing Loss: A Retrospective Cohort Study. The Journal of International Advanced Otology, 17(6), 514–519. https://doi.org/10.5152/iao.2021.21169
Jastreboff P. J. (2015). 25 years of tinnitus retraining therapy. HNO, 63(4), 307–311. https://doi.org/10.1007/s00106-014-2979-1
Jin, I. K., Choi, S. J., Yoo, J., Jeong, S., Heo, S., & Oh, H. (2021). Effects of Tinnitus Sound Therapy Determined Using Subjective Measurements. Journal of the American Academy of Audiology, 32(4), 212–218. https://doi.org/10.1055/s-0041-1722984
Wang, J., Ding, J., Song, J., Hu, L., Cong, N., & Han, Z. (2022). A Prospective Study of the Effect of Tinnitus Sound Matching Degree on the Efficacy of Customized Sound Therapy in Patients with Chronic Tinnitus. ORL; Journal for Oto-Rhino-Laryngology and its Related Specialties, 84(3), 229–237. https://doi.org/10.1159/000517631
Yale Medicine. (2023, May 30). Tinnitus. Yale Medicine. https://www.yalemedicine.org/conditions/tinnitus