Study of sequential emotional dynamics: a LSDVG approach with reduced electrode configuration

Area of Science:

• Neuroscience
• Computer Science
• Signal Processing

Background:

• Electroencephalogram (EEG) based emotion recognition is a reliable alternative to traditional methods.
• The impact of emotional stimuli history on current emotional response is under-investigated in EEG research.

Purpose of the Study:

• To investigate how the history of emotional stimuli influences current emotional responses using EEG.
• To develop a computationally efficient method for emotion recognition.

Main Methods:

• Preprocessing EEG data using local standard deviation (LSD).
• Transforming preprocessed data into a visibility graph (VG).
• Calculating VG topological properties (modularity, communities, density, degree, entropy) for classification.

Main Results:

• Highest classification accuracy was achieved with novel emotional sequences.
• Repetitive emotional stimuli led to decreased accuracy due to emotional familiarity.
• The LSD-VG approach achieved high accuracy using only 12 EEG channels.

Conclusions:

• The LSD-VG approach is computationally efficient for emotion recognition.
• Emotional history significantly influences current emotional brain responses.
• This method aids in understanding the brain's adaptation to repeated emotional stimuli.