Investigating EEG Patterns for Dual-Stimuli Induced Human Fear Emotional State

Area of Science:

• Neuroscience
• Affective Computing
• Biomedical Engineering

Background:

• Current electroencephalography (EEG) emotion recognition predominantly uses visual or auditory stimuli.
• Limited research exists on self-induced emotions and the comparison of different stimuli for eliciting the same emotion.
• The existence of subject-independent correlations across different emotion-eliciting stimuli remains largely unexplored.

Purpose of the Study:

• To introduce and validate a dual-modality emotion elicitation paradigm.
• To investigate if emotions, specifically fear, can be classified when induced by different stimuli (self-induced imagery vs. audio/video clips).
• To assess subject-independent correlations in EEG signals for fear emotions evoked by distinct stimuli.

Main Methods:

• A novel dual-modality paradigm was employed for emotion induction.
• Common Spatial Pattern (CSP) was used for feature extraction from EEG data.
• Linear Discriminant Analysis (LDA) was applied for emotion classification, with spectral analysis for correlation investigation.

Main Results:

• A classification method based on CSP and LDA was proposed and applied to fear emotions.
• The study identified critical EEG channels using spatial filter weights.
• Performance was enhanced by comparing CSP with other regularized techniques.

Conclusions:

• This research presents a novel approach to classifying emotions induced by different stimuli using EEG.
• It offers the first known assessment of EEG correlations between self-induced and video-induced emotions.
• The findings contribute to understanding subject-independent patterns in brain activity during emotional responses.