Task-Specific Rapid Auditory Perceptual Learning in Adult Cochlear Implant Recipients: What Could It Mean for Speech...

Area of Science:

• Auditory Neuroscience
• Speech Perception
• Rehabilitation Engineering

Background:

• Speech recognition varies significantly among cochlear implant (CI) recipients, especially in noisy environments.
• While demographic, audiological, and cognitive factors explain some variability, a significant portion remains unexplained.
• Rapid auditory perceptual learning is hypothesized to account for unique variance in speech recognition for individuals with and without hearing loss.

Purpose of the Study:

• To investigate if adult CI recipients exhibit rapid auditory perceptual learning.
• To determine if this learning contributes uniquely to speech recognition in challenging conditions (fast speech, speech in noise).
• To focus on the early adaptation phase of task-specific rapid auditory perceptual learning.

Main Methods:

• Thirty-six adult CI recipients completed speech recognition tests (noise, babble, fast speech) and cognitive assessments.
• A rapid auditory perceptual learning task using time-compressed speech was administered.
• Generalized mixed models were used to analyze speech recognition accuracy, accounting for demographic, audiological, cognitive factors, and rapid learning.

Main Results:

• Most CI recipients demonstrated early task-specific rapid auditory perceptual learning within the first 20 sentences.
• This learning significantly contributed to recognizing fast speech in quiet and speech in noise.
• A 1 SD increase in learning rate correlated with a ~52% increase in odds of recognizing fast speech and ~19-28% for speech in noise, after controlling for other factors.

Conclusions:

• Early rapid auditory perceptual learning in CI recipients accounts for individual differences in speech recognition, particularly in adverse conditions.
• The ability to rapidly adapt to acoustic challenges is a crucial factor for successful CI outcomes.
• Higher cognitive resources and faster learning rates are associated with better speech recognition in CI users.