Identification of Salient Brain Regions for Anxiety Disorders Using Nonlinear EEG Feature Analysis.

Area of Science:

• Neuroscience
• Computational Psychiatry
• Biomedical Engineering

Background:

• Anxiety disorders represent a significant global health burden.
• Accurate diagnosis and identification of neural correlates are crucial for effective treatment.
• Current diagnostic methods may lack objective biomarkers for anxiety disorders.

Purpose of the Study:

• To develop and validate a novel approach for identifying salient brain regions in anxiety disorders using nonlinear electroencephalography (EEG) features.
• To interpret the discriminative ability of these nonlinear EEG features between patients with anxiety disorders and healthy controls.
• To enhance the interpretability of machine learning models for identifying relevant brain regions in anxiety disorders.

Main Methods:

• Advanced EEG preprocessing and artifact correction techniques were employed.
• Nonlinear feature extraction was performed using conditional permutation entropy.
• Interpretable machine learning models were utilized to identify relevant electrodes and brain regions.

Main Results:

• Extracted nonlinear EEG features demonstrated statistically significant differences between anxiety disorder patients and healthy controls (T-tests: p = 1.05e-10; Mann-Whitney U tests: p = 2.65e-11).
• The features exhibited high discriminative ability, confirmed by robust statistical significance.
• Classification results successfully guided the identification of relevant electrodes, enhancing feature interpretability.

Conclusions:

• The proposed method effectively identifies brain regions critical for discriminating anxiety disorders.
• This approach offers a promising tool for objective diagnosis of anxiety disorders.
• Findings pave the way for targeted interventions and improved clinical outcomes in anxiety disorder management.