Generative AI for suppressing tinnitus

Tinnitus is a phantom perception, where a sound is perceived by the patient that cannot be objectively perceived by other observers [1]. This can have debilitating effects on tinnitus sufferers, leading to anxiety, depression, and insomnia, amongst other symptoms [2].

Almost 10% of people suffer from chronic tinnitus [3], where they are constantly exposed to an annoying sound or noise. Just recently, a new treatment method has been developed at the bioelectronics department at the TU Delft that utilizes neuromodulation to treat tinnitus. It combines electrical stimulation of the vagus nerve with acoustic stimulation using a tinnitus-matched sound. The aim is to teach the brain new associations so that it labels the tinnitus sound as being unimportant and will suppress the tinnitus. A startup and spin-off of the TU Delft has taken up the challenge to develop the bimodal stimulator further.

For a successful treatment, it is very important to accurately mimic the tinnitus sound that a patient perceives. The tinnitus sound that people perceive can vary greatly between individuals, ranging from a single tone to a complex sound spanning a large frequency spectrum. Sounds may even vary slightly in pitch or loudness over time. The complexity of possible tinnitus sounds makes tinnitus-matching difficult. This is complicated by the fact that people generally have difficulty describing their tinnitus sound, especially when they lack any musical experience. For example, the tinnitus-matched sound might be a whole octave off from the actual tinnitus sound.

Current methods are not able to accurately match the tinnitus sound. Most of them only focus on matching a single frequency instead of matching more complex tones. Furthermore, methods do not take into account the user interaction, thereby complicating the matching process and making it less likely that the tinnitus is accurately matched. The lack of a good tinnitus-matching method lowers the probability of success of the proposed treatment.

[1] D. De Ridder e.a., ‘Tinnitus and tinnitus disorder: Theoretical and operational definitions (an international multidisciplinary proposal)’, Prog Brain Res, vol. 260, pp. 1-25, 2021, doi: 10.1016/bs.pbr.2020.12.002.

[2] B. Langguth, ‘A review of tinnitus symptoms beyond “ringing in the ears”: a call to action’, Curr Med Res Opin, vol. 27, nr. 8, pp. 1635-1643, aug. 2011, doi: 10.1185/03007995.2011.595781.

[3] C. M. Jarach e.a., ‘Global Prevalence and Incidence of Tinnitus: A Systematic Review and Meta-analysis’, JAMA Neurology, vol. 79, nr. 9, pp. 888-900, sep. 2022, doi: 10.1001/jamaneurol.2022.2189.

Assignment

This MSc project aims to improve the current tinnitus-matching methods available through for example a literature study and / or the design and testing of a prototype. The project can still be shaped according to your interests or expertise. A literature study on the currently used methods for tinnitus-matching can be conducted. The design of a prototype may consist of an algorithm, application, or device using AI such as generative AI. Also, implementing a UI and designing an appropriate user experience can be important aspects of a tinnitus-matching method.

This project is in collaboration with Prof. Wouter Serdijn and Jonathan Kneepkens (TU Delft), Dr. Christos Strydis (Erasmus MC and TU Delft) and Prof. Dr. Dirk De Ridder (University of Otago, New Zealand).

Contact

Last modified: 2024-09-06