Towards artifact-free auditory evoked potentials in cochlear implant users [Elektronische Ressource] / Filipa Alexandra Campos Viola. Betreuer: Stefan Debener

Towards artifact-free auditory evoked
potentials in cochlear implant users


Der Fakultät für Mathematik und Naturwissenschaften
der Carl von Ossietkzy Universität Oldenburg
angenommene Dissertation zur Erlangung
des Grades und Titels eines


Doctor rerum naturalium (Dr. rer. nat.)



von Frau Filipa Alexandra Campos Viola
geboren am 01.01.1982 in Beja, Portugal.




Gutachter: Prof. Dr. Stefan Debener

Zweitgutachter: Prof. Dr. Christoph Herrmann

Tag der Disputation: 30.11.2011




Erklärung nach § 10 Abs. 2 der Promotionsordnung der Fakultät für Mathematik
und Naturwissenschaften der Carl von Ossietzky Universität Oldenburg vom
11. Dezember 2003

Hiermit erkäre ich, dass ich die Arbeit selbständig verfasst und nur die angegeben
Hilfsmittle benuzt habe. Die Dissertation ist bereits in Gänze veröffentlicht.


Oldenburg, den ___________ ______________________________________
Unterschrift




Erklärung nach § 10 Abs. 3 der Promotionsordnung der Fakultät für Mathematik
und Naturwissenschaften der Carl von Ossietzky Universität Oldenburg vom
11. Dezember 2003

Hiermit erkäre ich, dass die Dissertation weder in ihrer Gesamtheit noch in Teilen einer
anderen wissenschaftlichen Hochschule zur Begutachtung in einem Promotionsverfahren
vorliegt oder vorgelegen hat.


Oldenburg, den ___________ ______________________________________
Unterschrift




Abstract
Cochlear implants (CI) are neural prostheses that mimic the function of the healthy
cochlea and deliver electrical stimulation to the auditory nerve, allowing individuals
suffering from sensorineural hearing loss to recover a large amount of hearing function.
Although CIs are regarded as one of the great achievements of modern medicine, the
outcomes after implantation are variable, and it is not well understood how the auditory
cortex adapts to the electrical stimulation delivered by the CI. An objective and non-
invasive method for assessing auditory rehabilitation after implantation is by using
electroencephalography (EEG) to measure auditory evoked potentials (AEPs). However
EEG signals consist of a mixture of an unknown number of brain and non-brain
contributions, the latter also called artifacts. The non-brain signals can be divided into
two categories: biological and non-biological artifacts. CI devices cause non-biological
electrical artifacts
These artifacts corrupt and mask the AEPs, since they are time-locked to auditory stimuli,
and therefore cannot be attenuated using standard techniques such as filtering or
averaging. A particularly promising technique to deal with CI artifacts is independent
component analysis (ICA). This technique can disentangle multi-channel EEG signals
into a number of artifacts and brain-related signals, also called independent components
(ICs). When using an ICA-based attenuation approach, the ICs related to artifacts can be
removed, and a corrected version of the original EEG signal can be obtained. However,
the identification and interpretation of ICs is time-consuming and involves subjective
decision making.
In Study 1 a tool tailored to identify ICs representing eye blinks, lateral movements and
heartbeat artifacts was developed and validated. The tool is based on the correlation of
ICA inverse weights, also called IC scalp maps, with a user-defined template map, thus it
was named CORRMAP. The performance of the tool was compared with the
performance of 11 raters from different laboratories familiar with ICA. The overlap
between ICs selected by CORRMAP and by the raters was substantial, providing
evidence that the tool offers an efficient way of attenuating these particular artifacts.
In Study 2 the effects of CI artifact attenuation on AEP quality were investigated in a
sample of 18 adult post-lingually deafened individuals, using different types of CIs. Here
ICs related to CI artifacts were selected by visual inspection and AEPs were
reconstructed. It was found that AEPs from CI users were systematically correlated with


age, indicating that individual differences were well preserved. CI users with large signal-
to-noise ratio AEPs were characterized by a significantly shorter duration of deafness.
The ability of ICA in attenuating the CI