| Market dynamics, technology spin-cycles, and affordability pressures are shaping development demands for medical devices. Model-based development offers a remedy for these pressures, with demonstrated ability to mitigate the complexity introduced. This dissertation will examine the program lifecycle for a representative new medical device, the Optical Cochlear Implant, and will identify high-leverage touchpoints for model-based development. As part of the study, we'll analyze gaps and possible mitigations in the field of hearing impairment, including a look at the current frontier of research in hearing aids and cochlear implants. The Optical Cochlear Implant system is a new device, which has the potential for an order of magnitude improvement in spectral fidelity over conventional devices. Its projected development timeline spans many years: from point testing in the laboratory, through approval process for limited human testing, with control gates for application to human candidates. In the journey from lab to human patient, design decisions for medical devices are often made in absence of understanding the system-level impacts those design decisions can have on the ultimate utility of the device. Since medical devices generally demonstrate intense coupling between the patient and outcome, faulty design decisions are often identified late in the device development when correction is much more expensive and could trigger a need for re-certification which would drive development timelines even longer. The foundation for model-based system development of Optical Cochlear Implants is laid with the models and simulations developed in this dissertation that extend laboratory measurements in order to evaluate intelligibility impacts. Specifically, the Cochlear Laser Transduction Model, a physics-based simulator models tonotopic specificity; the Optical Cochlear Implant Simulator (OCIS), extends the laboratory measurements via simulation using the results of the Cochlear Laser Transduction Model; and an off-the-shelf speech recognition tool, CMU SPHINX models the human patient impacts via intelligibility assessments. Results from the system level modeling and simulation approach indicate that significant intelligibility improvements could be realized with the optical cochlear implant, with potential to offer upwards of 40 channels of spectral resolution compared to 10 channels with electrode-based implants. The resulting architecture and models allow investigation of device to outcome and support design and development of Optical Cochlear Implants. This dissertation includes analysis of the potential cost savings that could accrue due to this model-based approach compared to a conventional development approach. |
