SCH: INT: Collaborative Research: A Framework for Optimizing Hearing Aids In Situ Based on Patient Feedback, Auditory Context, and Audiologist Input

到2030年，20%的美国人将达到65岁或以上，其中35% - 50%的人将报告患有与年龄相关的听力障碍，主要使用助听器（HA）治疗。经常使用HAs已被证明可以改善沟通，避免听力损失的负面影响，包括增加社会孤立、抑郁和无法工作、旅行或身体活动的风险。过去的研究表明，许多人不经常戴has，因为他们对现实生活中的表现不满意。现有调优HA方法的一个基本限制是，它们没有针对个人需求进行定制，这通常会导致性能不令人满意。作为该项目的一部分，一个由计算机科学家、工程师和听力学家组成的团队将根据患者的个人需求开发调整人工耳蜗的新方法。该项目基于一种方法，即可以根据现场收集的患者反馈来确定更好的HA配置。该项目包括开发两个优化ha配置的系统。第一个系统将通过改进从患者那里获得反馈的方式，结合听觉环境信息，并将其转化为HA配置调整，从而提高传统调谐过程的效率。第二个系统将根据患者反馈和听觉环境信息自动调整HA的配置，而不需要听力学家进行调整。该提案的智力价值包括机器学习的进步，这是对具有众多参数（如HAs）的复杂系统的性能建模所必需的，并根据从患者那里获得的反馈调整其参数。此外，该项目将通过开发运行高可用性优化算法的技术，推动嵌入式系统的发展，使其成为由高可用性、移动电话和云服务组成的多层系统的一部分。预期本研究将使患者更积极参与听力护理，提高医管局的满意度，并丰富他们的社会互动。该奖项反映了美国国家科学基金会的法定使命，并通过使用基金会的知识价值和更广泛的影响审查标准进行评估，被认为值得支持。

Twenty percent of Americans will be 65 years or older by 2030 out of which 35 - 50% will report having age-related hearing impairment that is treated primarily with hearing aids (HA). Regular use of HAs has been shown to improve communication and avoid the negative effects of hearing loss that include an increased risk of social isolation, depression, and inability to work, travel, or be physically active. Past research has shown that many people do not wear their HAs regularly, as they are unsatisfied with the performance in the real world. A fundamental limitation of existing methods for tuning HA is that they are not tailored to individual needs, which often leads to unsatisfactory performance. As part of this project, a team of computer scientists, engineers, and audiologists will develop new methods for tuning HAs that are based on the individual needs of a patient.The project is based on the approach that better HA configurations may be identified based on feedback from the patient that is collected in the moment and in situ. The project includes development of two systems for optimizing the configuration of HAs. The first system will improve the efficacy of the traditional tuning process by improving how feedback is obtained from patients, combined with auditory context information, and translated into HA configuration adjustments. The second system will automatically adapt the configuration of a HA based on patient feedback and auditory context information without requiring an audiologist to perform adjustments. The intellectual merit of this proposal includes the advances in machine learning necessary to model the performance of complex systems that have numerous parameters such as HAs and tuning their parameters based on feedback obtained from patients. Additionally, the project will advance the state-of-the-art in embedded systems by developing techniques to run HA optimization algorithms as part of a multi-tier system composed of HAs, mobile phones, and cloud services. It is anticipated that the proposed research will empower patients to become more involved in their hearing care, improve HA satisfaction, and enrich their social interactions.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.