Cochlear prosthesis is widely accepted as the most effective clinical intervention to restore auditory function of individuals with profound hearing loss. Although state-of-the-art CIs provide a high level of speech comprehension and aural communication ability to a majority of implant recipients, there remains a major gap between performance levels of CI users and normal hearing individuals, especially in real-life noisy environments. This gap in performance in part can be attributed to limitations in both sound coding and electrical stimulation strategies, and partially due to the limited ability to explore potentially new advanced algorithms with current CI users in the field. Several methods have been proposed over the years to address this shortcoming; however, most have been restricted to laboratory research. This is primarily due to the unavailability of portable sound processing platforms that can 1) implement computationally-intensive sound processing schemes and 2) assess them chronically in real naturalistic environments. Clinical processors/platforms are neither powerful, nor flexible to meet the growing scientific needs of the research community. We propose a multi-center research effort to investigate three complementary sound processing strategies (Aims 1 ? 3), which will be made possible through the proposed research platform (Aim 4). First, we will develop and test the effectiveness of two new families of front-end speech processing algorithms (Aim 1), both of which are inspired by speech production/perception physiology and aim to enhance the speech signal from competing background noise. The potential benefit of these algorithms in real-life acoustic environments will be assessed by conducting take-home trials using the portable research platform. Next, we will investigate the potential benefits of real-time user-specified adjustments to frequency allocation and stimulation rate adjustments on speech perception and sound quality in naturalistic environments (Aim 2).
In Aim 3, we will investigate the effectiveness of speech processing strategies that deliver synchronized electrical stimulation to bilateral CIs. Specifically, we aim to test differences in ITD discrimination, sound localization, and segregation of speech in noise with and without synchronized bilateral stimulation. These studies will be done using the existing prototype of the platform, CCi-MOBILE. As a next step we propose to develop a next-generation CCi- MOBILE-2 platform - a flexible, open-source, portable sound processing platform that will allow easy implementation of research ideas as well as long-term assessments of algorithms in real-life acoustic environments (Aim 4). This one-of-a-kind research platform will be orders of magnitude more flexible and computationally powerful than existing clinical processors and will aid in bridging scientific research with commercial applications. The CCi-MOBILE-2 platform will be shared with the CI research community free of cost using an open source model. The experiments listed here represent a mere subset of the potential groundbreaking advancements that will be made possible by the existence of the proposed research platform. The ability to perform real-life chronic speech assessments will open new frontiers for scientific exploration and will result in a paradigm shift in how speech processing/perception research is carried out in the cochlear implant field. Advancements from Aims 1-3 will show clear examples of how to transition scientific and algorithmic advancements to field testing with the CCi-MOBILE-2 (Aim 4), thus giving operational examples on how to leverage the research platform for other research laboratories.
The proposed multi-center research effort seeks to explore the novel research directions made possible by a flexible, powerful, portable speech processor for cochlear implants based on commercially available smartphones/tablets. This research platform will be significantly more flexible and computationally powerful than clinical devices, and will enable implementation and long-term evaluation of advanced signal processing algorithms in naturalistic, diverse acoustic environments. It will be combined with take-home assessments of novel front-end speech enhancement/processing strategies that are inspired by human speech physiology (Aim 1), on-the-fly patient-specific sound processing adjustments (Aim 2), and stimulation strategies that provide enhanced binaural acoustic cues (Aim 3). Further development of the research platform is also planned (Aim 4), intended as an open-source contribution to the cochlear implant field.
