Effect of extreme adaptive frequency compression in bimodal listeners on sound localization and speech perception

Objectives: This study aimed to improve access to high-frequency interaural level differences (ILD), by
applying extreme frequency compression (FC) in the hearing aid (HA) of 13 bimodal listeners, using a
cochlear implant (CI) and conventional HA in opposite ears.
Design: An experimental signal-adaptive frequency-lowering algorithm was tested, compressing frequencies
above 160 Hz into the individual audible range of residual hearing, but only for consonants (adaptive FC),
thus protecting vowel formants, with the aim to preserve speech perception. In a cross-over design with at
least 5 weeks of acclimatization between sessions, bimodal performance with and without adaptive FC
was compared for horizontal sound localization, speech understanding in quiet and in noise, and vowel,
consonant and voice-pitch perception.
Results: On average, adaptive FC did not significantly affect any of the test results. Yet, two subjects who
were fitted with a relatively weak frequency compression ratio, showed improved horizontal sound
localization. After the study, four subjects preferred adaptive FC, four preferred standard frequency
mapping, and four had no preference. Noteworthy, the subjects preferring adaptive FC were those with
best performance on all tasks, both with and without adaptive FC.
Conclusion: On a group level, extreme adaptive FC did not change sound localization and speech
understanding in bimodal listeners. Possible reasons are too strong compression ratios, insufficient
residual hearing or that the adaptive switching, although preserving vowel perception, may have been
ineffective to produce consistent ILD cues. Individual results suggested that two subjects were able to
integrate the frequency-compressed HA input with that of the CI, and benefitted from enhanced binaural
cues for horizontal sound localization.