Abstract
Objective: The fundamental frequency modulation (F0mod) sound processing strategy was developed to improve pitch perception with cochlear implants. In previous work it has been shown to improve performance in a number of pitch-related tasks such as pitch ranking, familiar melody identification, and Mandarin Chinese tone identification. The objective of the current study was to compare speech perception with F0mod and the standard clinical advanced combination encoder (ACE) strategy. Study sample: Seven cochlear-implant listeners were recruited from the clinical population of the University Hospital Leuven. Design: F0mod was implemented on a real-time system. Speech recognition in quiet and noise was measured for seven cochlear-implant listeners, comparing F0mod with ACE, using three different Dutch-language speech materials. Additionally the F0 estimator used was evaluated physically, and pitch ranking performance was compared between F0mod and ACE. Results: Immediately after switch-on of the F0mod strategy, speech recognition in quiet and noise were similar for ACE and F0mod, for four out of seven listeners. The remaining three listeners were subjected to a short training protocol with F0mod, after which their performance was reassessed, and a significant improvement was found. Conclusions: As F0mod improves pitch perception, for the seven subjects tested it did not interfere with speech recognition in quiet and noise, and has a low computational complexity, it seems promising for implementation in a clinical sound processor.
Acknowledgements
We are grateful to our test subjects for their inexhaustible and enthusiastic collaboration. We thank the staff of the audiology team of the University Hospital Leuven for their support. This study was performed as part of IWT Flanders project number 110722 in cooperation with Cochlear Ltd.
Declaration of interest: The authors report no conflicts of interest.
Notes
1 The frequency contours were obtained by using the following Praat command: To pitch (ac)... 0.01 75 15 no 0.01 0.45 0.01 0.35 0.14 400.