Author:
Migliorini Enrico,Wasmann Jan-Willem A.,Philpott Nikki,van Dijk Bastiaan,Philips Birgit,Huinck Wendy
Abstract
AbstractPurposeCurrent clinical measures used in cochlear implantation (CI) provide a broader view of speech recognition ability at word-level, often missing granular details contained at phoneme-level that may be valuable for CI mapping. This study evaluates how outcomes of Phoneme Recognition in Quiet tests (PRQ) differ from those of more commonly used word recognition tests (CVC) and outlines how these tests may be useful for different purposes in clinical adult CI care.MethodsAs part of the AuDiET (Auditory Diagnostics and Error-based Treatment) study, 23 adult postlingually deafened unilateral CI users underwent a battery of tests, including both PRQ and CVC tests. Their results were compared at the phoneme level, including an evaluation of fitness and error dispersion.ResultsPRQ had a significantly lower accuracy and fitness than CVC. The error patterns also tended to be less random and more systematic. Fitness correlated strongly and positively with accuracy, while error dispersion negatively correlated with accuracy.ConclusionThere are clear differences between PRQ and CVC outcomes in absolute accuracy and error distribution. Comparing these tests might provide clinicians with more granular insights into which areas/phonemes to target during mapping, to achieve optimal speech recognition.
Publisher
Cold Spring Harbor Laboratory