RESUMO
The importance of simulating patient behavior for medical assessment training has grown in recent decades due to the increasing variety of simulation tools, including standardized/simulated patients, humanoid and android robot-patients. Yet, there is still a need for improvement of current android robot-patients to accurately simulate patient behavior, among which taking into account their hearing loss is of particular importance. This paper is the first to consider hearing loss simulation in an android robot-patient and its results provide valuable insights for future developments. For this purpose, an open-source dataset of audio data and audiograms from human listeners was used to simulate the effect of hearing loss on an automatic speech recognition (ASR) system. The performance of the system was evaluated in terms of both word error rate (WER) and word information preserved (WIP). Comparing different ASR models commonly used in robotics, it appears that the model size alone is insufficient to predict ASR performance in presence of simulated hearing loss. However, though absolute values of WER and WIP do not predict the intelligibility for human listeners, they do highly correlate with it and thus could be used, for example, to compare the performance of hearing aid algorithms.
RESUMO
Today, medical technology manufacturers enter the service market through the development of digital service innovations. In the field of audiology, these developments increasingly shift the service capacities from audiologists to manufacturers and technical systems. However, the technology-driven developments of manufacturers lack acceptance of hearing device users and undermine the important role of audiologists within the service provision. By following a user-centered design approach in order to deal with the technological and social challenges of disruptive services, we aim to develop service innovations on an integrated service platform in the field of tele-audiology. To ensure the acceptance of technology-driven service innovations among hearing device users and audiologists, we systematically integrated these actors in a participatory innovation process. With qualitative and quantitative data we identified several requirements and preferences for different service innovations in the field of tele-audiology. According to the preferences of the different actors, we proposed a service platform approach based on a connected hearing device in three pillars of application: 1) one-to-one (1:1) service innovations based on a remote fitting concept directly improve the availability of services offered by audiologists without being physically present. Based on this, 2) one-to-many (1:N) service innovations allow the use of the connected hearing device as an indirect data source for training a machine learning algorithm that empowers users through the automation of service processes. A centralized server system collects the data and performs the training of this algorithm. The optimized algorithm is provided to the connected hearing devices to perform automatic acoustic scene classification. This in turn allows optimization of the hearing devices within each acoustic scene. After the user-centered development of the different service innovations which are designed to converge on an integrated service platform, we experimentally evaluated the functionality and applicability of the system as well as the associated role models between the technical system, the hearing device users and audiologists. As a future outlook, we show potentials to use the connected hearing device for 3) cross-industry (N:M) service innovations in contexts outside the healthcare domain and give practical implications for the market launch of successful service innovations in the field of tele-audiology.