Investigation of Objective Neurophysiological Measures of Auditory Temporal Processing Abilities and Their Applicability in Cochlear Implant Rehabilitation
Citation:
WAECHTER, SASKIA MARLEEN, Investigation of Objective Neurophysiological Measures of Auditory Temporal Processing Abilities and Their Applicability in Cochlear Implant Rehabilitation, Trinity College Dublin.School of Engineering, 2019Download Item:
Abstract:
Speech is the most important form of human communication. For people affected by disabling hearing loss, speech information may become unavailable causing social isolation and cognitive decline. People with mild to profound sensorineural hearing loss may benefit from cochlear implantation which can bypass the impaired auditory pathway and may restore functional hearing through direct electrical stimulation. The fitting process, which refers to the continuous adjustment of software parameters of the cochlear implant (CI) after implantation, is an important part of the success of clinical rehabilitation procedures. Current fitting approaches are strongly reliant on subjective feedback, which can be unreliable, e.g. in individuals with test anxiety, or missing in cohorts such as infants and people with cognitive impairments. For children who are born profoundly deaf, it is imperative that they receive a CI as early as possible in life to achieve optimal rehabilitation outcomes, which has led to a large increase of cochlear implantation at a young age. Therefore, there is an urgent clinical need for objective methods to assess auditory processing abilities after implantation. Such objective metrics can not only guide the clinical fitting procedures, but also provide additional information to aid expectation management particularly for parents of infants who receive a CI.
In the literature, potential objective metrics have been evaluated for spectral as well as temporal auditory processing. Both, temporal and spectral acoustic cues, contribute to the intelligibility and recognisability of sounds and particularly of speech. A multitude of research studies has assessed objective metrics of spectral resolution, however, fewer research studies have investigated objective metrics of temporal auditory processing, which is of particular importance for speech perception. In this thesis, objective neurophysiological measures representative of auditory discrimination abilities were explored for two different aspects of temporal auditory processing, and their applicability in a clinical cohort was investigated. The temporal features under investigation were amplitude modulation (AM) detection abilities (Chapter 3 - Chapter 5) and discrimination abilities based on temporal fine structure (TFS) cues (Chapter 6), both of which play an important role in speech recognition.
The first study focused on the methodology of how to derive individual neural thresholds of auditory discrimination abilities from electroencephalography (EEG) data. Consecutively, the relationship between these neural thresholds and corresponding behavioural thresholds was analysed by means of correlation analysis to assess the informational value of the neural threshold estimates. Successful neural threshold estimation and significant correlations with behavioural thresholds provided support for the expansion to a clinical cohort in Study 2. This study addressed another important research question regarding the applicability of the novel approach in a clinical CI user cohort. Moreover, signal processing methods were proposed to automatically (Case Study) or manually (Study 2) reduce the electrical artefact in EEG data which arises from CI stimulation. Significant correlations between behavioural and neural thresholds obtained in Study 2 corroborate findings from Study 1 and provide evidence for the feasibility of employing such an objective metric in a clinical CI user cohort.
The third study as well as multiple accompanying pilot studies have explored tone discrimination based on two types of TFS cues and whether cortical neural change detection measures reflect behavioural discrimination abilities. Findings showed that change detection based on TFS cues is encoded in cortical neural measures, if tones can be confidently distinguished. However, measured amplitudes were low in comparison to studies which employ acoustic change features based on spectral or intensity cues. Moreover, for one of the two types of assessed TFS cues, precise stimulus replication based on information in the literature was unsuccessful despite further correspondence with authors, highlighting the need for more thorough descriptions of employed methods in Journal publications.
In conclusion, the original contribution of this thesis to the literature has multiple facets and includes the novel application of widely researched neurophysiological measures of auditory change detection to acoustic features that have previously not been assessed with this approach, such as TFS processing and AM detection for varying AM depths. Furthermore, novel signal processing methodologies have been proposed to enhance neural signal analysis on an individual level. This included the development of a new methodology to quantify neurophysiological change responses, to aid the estimation of individual neural discrimination thresholds. These thresholds showed significant correlations with behavioural discrimination thresholds in a normal-hearing cohort as well as in a clinical CI cohort, and thus, may enhance the objective assessment of auditory processing abilities in clinical environments in the future.
Sponsor
Grant Number
IBM Research Award (part funded)
Cochlear Ltd (part funded)
Trinity Centre for Biomedical Engineering (part funded)
Author's Homepage:
https://tcdlocalportal.tcd.ie/pls/EnterApex/f?p=800:71:0::::P71_USERNAME:WAECHTESDescription:
APPROVED
Author: WAECHTER, SASKIA MARLEEN
Advisor:
Reilly, RichardPublisher:
Trinity College Dublin. School of Engineering. Discipline of Electronic & Elect. EngineeringType of material:
ThesisCollections
Availability:
Full text availableMetadata
Show full item recordLicences: