Dynamic adjustment of cochlear implants to improve speech comprehension

This project aims to provide a novel aproach to cochlear implant speech processing. Former speech processing strategies try to present the frequencies to the patient as natural as possible. The present approach uses frequency transposition and compression to shift the incoming signal into the remaining narrow frequency band based on the dominant frequency of the signal. Due to the better perception of key speech features, the understanding will likely improve.

• sketch out the idea
• develop a frequency transpositioning algorithm using digital signal processing (DSP) techniques
• test approach in a study (either with normal hearing people or, if available, acquire support and test with CI users)
• implement algorithm as module in embedded system to test hardware requirements and evaluate resource usage
• improve resource usage and quality

• developed new algorithm for enhanced time-stretching used by frequency transposition algorithms
• conception of the DSP-based algorithm: Adaptive Pitch Transposition
• prototypical implementation of the algorithm
• implementation of a live-output capable version of the prototype

Peer-reviewed journal papers:

STRUWE, K., 2017. APT: Enhanced Speech Comprehension Through Adaptive Pitch Transposition in Cochlear Implants. Springer International Publishing. eHealth 360°, 181, pp. 224-228. DOI: 10.1007/978-3-319-49655-9_30.

• Prof. Dr.-Ing. Ralf Salomon
• Prof. Dr. Olaf Wolkenhauer