Presentation
Search Abstracts | Symposia | Slide Sessions | Poster Sessions | Lightning Talks
ERP signals during articulation: should we remove AEP?
Poster D53 in Poster Session D, Wednesday, October 25, 4:45 - 6:30 pm CEST, Espace Vieux-Port
Michael DePretto1, Marina Laganaro1; 1University of Geneva
In order to study speech planning of multiple syllables/words, it may be necessary to analyse the EEG event-related potentials (ERPs) signal during articulation. However, producing speech involves the overlapping of motor- and auditory/language- related processes. When analysing ERPs during articulation, muscle artefacts but also auditory evoked potentials (AEPs) may mask components related to the encoding of upcoming speech. Muscle artefacts can be removed with appropriate filtering and averaging across trials, or with the use of specific cleaning algorithms. Surprisingly, the processing of AEPs in the study of articulatory production has received, to our knowledge, little to no attention. Thus, the current project focused on distinguishing the signal related to speech planning and programming from the signal related to auditory processing during speaking. In a delayed production task, participants were asked to overtly produce two syllable non-words when a question mark appeared on the screen. On certain trials, no question mark appeared and they had to remain silent. In a second part, they had to listen to the recording of their own productions. In order to keep them focused on the task, they had to report occasional productions from another speaker. While they completed the task, we recorded the EEG signal from 128 channels. Here we report the preliminary results from seven young adults. We will increase our sample size until the meeting. For the analysis, we extracted ERPs aligned to the vocal onset in a time window running from 300 ms pre- to 200 ms post-vocal onset. After subtracting the ERPs associated to the listening signal from the ERPs of the production task, we compared the production average signal as recorded to the production average signal after subtraction. A mictrostate analysis showed that around 100-200 ms following vocal onset, the subtraction permitted to highlight a central positivity that was hidden by the AEP N1 component. The presence of this specific topographic map was accompanied by an increase in the global field power (GFP), indicating that the AEP N1 cancelled the central positivity. The rest of the signal in the time window of interest showed the same topographic maps and similar GFP amplitudes before and after subtraction. This result offers promising perspectives for the analysis of speech production related processes during articulation.
Topic Areas: Speech Motor Control, Methods