Brain responses to the McGurk effect in aphasia

The process of speech perception consists of different steps (Ellis & Young,
1988): auditory analysis, lexical processing and semantic processing, which can
be disturbed independently from each other. The auditory analysis is influenced
by lip-reading information. This can be in a helpful way (in noise or with
aphasia) as well as in a misleading way (the McGurk effect, in which the
lipmovement is different from the heard sound, leading to the perception of a
different sound). Influences of both, facilitating and misleading lip-reading,
were shown in a group of aphasic subjects in our previous behavioral research
(METc-Nr. 08.062; Hessler, Jonkers & Bastiaanse, 2009), indicating that
audiovisual integration takes place. It is however yet unclear how this
integration is accomplished in the brain. This will be subject to investigation
in the proposed study.
In order to answer this question event-related-potentials (ERPs) will be
recorded while measuring EEGs. The component aimed at (Mismatch Negativity -
MMN) has been found for this kind of experiment with healthy subjects before
and also in a different context with aphasic subjects.

Within this project we want to establish whether there
(1) are differences in brain responses to the McGurk effect between aphasic and
non-brain damaged control subjects
(2) is a difference between the behavioural performance of the aphasic subjects
and their brain responses
(3) the brain integrates lip-reading information even if the subject does not
do so consciously.

The general aims are therefore to refine theories about speech processing and
to provide a better basis for the development of diagnostic tools and treatment
programs.

*Event-related Potentials*(ERPs) will be measured during the processing of
audio-visual, auditory or visual stimuli. A sequence of stimuli (short
syllables formed by a consonant and a vowel) will be presented, consisting of
*standards* (occurring 80%) and two or three different deviants (occurring
7-10% each). The task is to press a button when encountering a deviant. This
task will be carried out in 4 different conditions: Pure tones, auditory only,
visual only and audiovisual (congruent and McGurk stimuli). In the pure tone,
auditory only and visual only there will be 800 standards and 100+100 deviants.
For the audiovisual condition, where there are three deviants, those will occur
100 times each, with 1200 standards (to establish the same ratio in all
conditions). All stimuli will be 800 ms long and will be presented with a
stimulus onset asynchrony of 1500 ms.

During the task electrical brain activity will be recorded with 64-channel EEG
(integrated in a cap). Four additional electrodes will be placed around the
eyes (EOG), two behind the ears and one on the sternum. To enlarge the EEG
signal an amplifier will be used (Twente Medical Systems BV, Hengelo)

There are no risks for the participants. The load is purely cognitive and
consist of maximally 2 times 3 hours. The participants will have to come to the
Neuroimaging Center in Groningen.