Investigating the applicability of single-channel electroencephalography in cochlear implant patients
Citation:Alejandro Lopez Valdes, 'Investigating the applicability of single-channel electroencephalography in cochlear implant patients', [thesis], Trinity College (Dublin, Ireland). Department of Mechanical and Manufacturing Engineering, 2017
Valdes, Alejandro Lopez_FullDocument_Final_Print.pdf (PDF) 12.82Mb
Hearing impairment affects over 5% of the population worldwide. Namely, about a quarter of a million adults in Ireland are estimated to have permanent hearing impairment due mainly to ageing and/or noise exposure. Cochlear implants (CI) can partially restore hearing in severely deafened patients by electrically stimulating the auditory nerve. Effective clinical assessment and adequate device programing is crucial for the rehabilitation of hearing after implantation. Current behavioural gold standards for assessing CI performance may fall short when assessing younger patient populations or difficult to test individuals. This thesis presents the results of investigating the applicability of objective electrophysiological metrics for assessing CI performance, introducing a new approach to reduce the effect of electrical artefact in CI electroencephalography recordings (EEG), deriving new neural based correlates of CI performance and evaluating them in a clinical environment. Chapter 3 presents the development of a three-stage CI artefact attenuation methodology. The aim of the chapter is to characterise the CI related artefact as seen in high resolution, single-channel, EEG recordings. The full representation of the artefact allowed for the development of an attenuation framework that succeeds in the extraction of the desired neural signal from the contaminated EEG data. The chapter details the origins of the CI artefact and discusses the application of single-channel EEG recordings in CI populations. Chapter 4 presents the development of an objective metric of spectral ripple discrimination based on single-channel EEG recordings and a mismatch paradigm. It is shown that neural estimates of spectral ripple discrimination correlate with traditional, behaviourally acquired, spectral ripple discrimination thresholds in a CI population. Due to the established relationship between spectral ripple discrimination and speech perception performance in CI users, the findings from this chapter suggest that CI performance could be objectively assessed via single-channel EEG. Chapter 5 investigates an alternative approach to objectively estimate spectral ripple discrimination in CI users. Exploring different EEG paradigms, this chapter provides a comparison of neural spectral ripple discrimination estimates as derived with a mismatch paradigm and an acoustic change (ACC) paradigm. The findings reported in this chapter suggest that, while it is possible to acquire neural spectral ripple discrimination thresholds via an ACC paradigm, the mismatch paradigm proved to be a more robust approach when estimating objective thresholds. Chapter 6 integrates the methods and tools introduced in the previous chapters for a longitudinal and clinical evaluation of spectral ripple discrimination and speech perception performance in newly implanted CI users. The aim of this chapter is to characterise the dynamics of the objective spectral ripple discrimination metric and its relationship with the dynamics of speech perception performance. This chapter presents advantages, challenges and limitations of the implementation of single-channel EEG recordings in a clinical environment. The results reported in this chapter provide supporting evidence that spectral ripple discrimination is a potential acute predictor of speech perception performance. Furthermore, EEG results suggest that neural estimates of spectral ripple discrimination may require longer maturation time when assessing CI performance. Chapter 7 investigates the development of an objective metric of temporal processing, specifically fine structure processing, in a CI population based on a mismatch paradigm and Schroeder-phase harmonic complexes. It is shown that the artefact attenuation methodology presented in Chapter 3 is suitable for use with temporal stimuli like the Schroeder-phase harmonic complexes. However, the adequate evaluation of an objective neural metric for temporal fine structure discrimination seemed to be hindered by the poor performance of this CI population. The single-channel EEG approach implemented in this thesis suffers from limited data dimensionality. However, its high temporal resolution allows for a complete characterisation of CI artefact. It has been shown that artefact free EEG responses can be retrieved from CI users under three different paradigms. The results presented in this thesis provide a valuable insight into the application of single-channel EEG for the clinical assessment of CI performance. Further research should be directed to the optimisation of the paradigms to be utilised in the clinic.
Programme for Research in Third-Level Institutions
European Regional Development fund
Author: Valdes, Alejandro Lopez
Advisor:Reilly, Richard B.
Publisher:Trinity College (Dublin, Ireland). Department of Mechanical and Manufacturing Engineering
Note:TARA (Trinity’s Access to Research Archive) has a robust takedown policy. Please contact us if you have any concerns: email@example.com
Type of material:thesis
Availability:Full text available