Time-compression thresholds for Mandarin sentences in normal-hearing and cochlear implant listeners

Area of Science:

• Auditory Neuroscience
• Speech Perception
• Cochlear Implants

Background:

• Faster speech can improve communication efficiency but may reduce intelligibility.
• Understanding speech processing limits is crucial for individuals with hearing impairments, particularly cochlear implant (CI) users.

Purpose of the Study:

• To quantify the maximum speech rate (time-compression thresholds or TCTs) for intelligible Mandarin sentences in normal hearing (NH) and CI listeners.
• To investigate the relationship between TCTs and speech reception thresholds in noise for CI users.

Main Methods:

• Employed an adaptive staircase procedure to measure 50%-word recognition thresholds for time-compressed Mandarin speech.
• Tested both original and CI-vocoded speech with NH subjects, and original speech with CI subjects.
• Speech was time-compressed without altering spectral characteristics, presented up to three times per trial.

Main Results:

• Median TCTs for NH listeners were approximately 16.7 syllables/s for normal speech, decreasing with fewer vocoder channels (11.8 and 8.6 syllables/s).
• CI listeners demonstrated significantly lower median TCTs, around 6.8 syllables/s.
• TCTs showed a strong correlation with speech reception thresholds in noise for CI listeners.

Conclusions:

• Normal hearing listeners process Mandarin sentences at rates up to 16.7 syllables/s, while CI users rarely exceed 10.0 syllables/s.
• These findings quantitatively define the upper limits of fast speech information processing for individuals using cochlear implants.