Different regressors for linear modelling of ElectroEncephaloGraphic recordings in visual and auditory tasks

• Published in: Proceedings (International Conference on Wearable and Implantable Body Sensor Networks : Print) pp. 260 - 265
• Main Author: Mugruza-Vassallo, Carlos
• Format: Conference Proceeding Journal Article
• Language: English
• Published: IEEE 01.06.2016
• Subjects: Analytical models; Auditory tasks; Brain models; Electrodes; Electroencephalography; Enterprise resource planning; Face; Modelling; Networks; Sensors; Visual; Visualization; Wearable
• ISSN: 2376-8894
• DOI: 10.1109/BSN.2016.7516270

The use of hierarchical linear modelling has been increasing in the last 5 years to analyze EEG data. Until now, no clear comparison on linear modelling in different modalities has been done. Therefore, specific differences observed in both visual and auditory paradigms were computed with linear modelling. The Coefficient of Determination through the explained variance (R2) in Linear Modelling was sought in visual and auditory modalities. ERP scalp series of time from 100 to 300 ms for the visual task and around 150 ms to 400 for the auditory task were also plotted. Although these paradigms use different regressors, both paradigms showed reliable R2 signatures across the participants and reliable ERP scalp maps. Results accounted for different magnitudes in greater R2 values for visual modality. Auditory R2 results appeared with a reliable linear modelling when compared with R2 studies in other subjects.