Predicting Speech Recognition Using the Speech Intelligibility Index and Other Variables for Cochlear Implant Users

Area of Science:

• Audiology
• Speech Science
• Biomedical Engineering

Background:

• The Speech Intelligibility Index (SII) is a valuable metric in audiology.
• Its application to cochlear implant (CI) users remains unexplored.
• Investigating SII's predictive power for CI users is crucial.

Purpose of the Study:

• To determine if the SII can effectively predict speech perception in individuals with CI.
• To explore the development of new SII models tailored for CI users.

Main Methods:

• Fifteen pre- and postlingually deafened adults with CI participated.
• Speech recognition was assessed using AzBio sentence lists.
• Psychoacoustic, cognitive, and demographic factors were analyzed alongside SII.

Main Results:

• Traditional SII calculations failed to predict speech performance in CI users due to hearing variability.
• New SII models, integrating demographic and perceptual-cognitive factors, significantly enhanced prediction accuracy.
• Aided audibility, duration of deafness, gap detection, and auditory digit span improved SII predictions.

Conclusions:

• Conventional SII models are inadequate for predicting speech perception in CI users.
• Demographic and perceptual-cognitive variables are essential for refining SII models for CI populations.
• Further research is needed to develop comprehensive CI-corrected SII models.