Feature Extraction to Identify Depression and Anxiety Based on EEG

Area of Science:

• Neuroscience
• Psychiatry
• Biomedical Engineering

Background:

• Mood disorders like depression and anxiety significantly impact individuals globally.
• Objective diagnostic tools are crucial for accurate identification and treatment of these conditions.
• Neurophysiological signals, such as EEG, offer a potential avenue for objective biomarker discovery.

Purpose of the Study:

• To investigate EEG signal characteristics as potential biomarkers for mood disorders.
• To explore differences in brain activity patterns between individuals with and without anxiety/depression.
• To assess the utility of physiological asymmetry as an indicator of mood disorders.

Main Methods:

• Analysis of EEG data from 119 young adults (18-24 years) performing a cognitive task.
• Classification of subjects into groups based on standard psychological assessments for anxiety and depression.
• Preprocessing of EEG signals, separation into frequency bands (beta, alpha, theta, delta), and extraction of complexity features (e.g., Higuchi Fractal Dimension, Approximate Entropy).
• Examination of hemispheric and topographical differences in EEG features using ANOVA II analysis.

Main Results:

• Significant differences (p<0.05) in several EEG features were observed between subjects with and without moderate to severe anxiety/depression.
• Evidence of physiological asymmetry in brain activity was found to be associated with higher scores on mood disorder assessments.
• Specific topographical regions showed distinct feature patterns differentiating affected from unaffected individuals.

Conclusions:

• EEG-derived physiological asymmetry shows promise as an objective indicator for mood and anxiety disorders.
• Understanding the neurophysiological underpinnings of mood disorders is critical for improving diagnostic accuracy.
• Further research into EEG biomarkers could lead to more objective diagnostic methods for psychiatric conditions.