Entwurf eines Hörtests zur Bestimmung des Comodulation Masking Release bei Cochlea Implantat-Trägern

Normal hearing (NH) listeners commonly perform better when
listening to speech in modulated than in steady state maskers. This
improvement is often labeled as a “release of masking”.
Cochlear-implant (CI) listeners generally don’t receive a masking
release in speech perception tests and the reasons are indistinct.
This study for the first time investigates the comodulation masking
release, an underlying psychoacoustic effect, in a flanking band
setup acoustically in CI users. The focus of the experiments was,
if CI users can use across-frequency cues to improve signal
detection in (co-)modulated interfering noise. The test bases upon
a forced choice paradigm to determine detection thresholds of a
sinusoidal signal in fluctuating maskers. The maskers consisted of
one or five amplitude modulated narrow noise-bands. The masking
noise-bands were either incoherently or coherently modulated
(comodulated). The difference of the detection threshold of the
sinusoidal signal in these two conditions is defined as
comodulation masking release (CMR). Two parameters of the masking
noise complex were varied: the bandwidth/modulation frequency and
the spectral alignment. 18 NH and 28 CI users participated in this
study. The measured detection thresholds in CI users were highly
reproducible. In contrast, the inter-individual standard deviation
was higher for CI users compared to NH. Nevertheless, CMR in CI
users was significant (p