Abnormal nonlinear features of EEG microstate sequence in obsessive-compulsive disorder.

Area of Science:

• Neuroscience
• Computational Psychiatry
• Biomarker Discovery

Background:

• Limited and inconsistent research exists on electroencephalography (EEG) microstates in obsessive-compulsive disorder (OCD).
• The nonlinear dynamics of EEG microstate sequences, crucial for brain information processing, remain unexplored in OCD.

Purpose of the Study:

• To investigate the nonlinear features of EEG microstate sequences in patients with OCD.
• To assess the potential of these nonlinear features as electrophysiological biomarkers for OCD detection.

Main Methods:

• Collected resting-state EEG data from 48 OCD patients and 48 healthy controls (HC).
• Analyzed EEG microstates to extract temporal parameters and nonlinear features (sample entropy, Lempel-Ziv complexity, Hurst index).
• Utilized machine learning models to classify OCD patients based on extracted features.

Main Results:

• OCD patients exhibited altered microstate durations (decreased A, B, C; increased D) and nonlinear features (increased sample entropy and Lempel-Ziv complexity; decreased Hurst index) compared to HC.
• Machine learning models achieved up to 85% classification accuracy using nonlinear features, outperforming models based on temporal parameters.
• Nonlinear features of EEG microstate sequences demonstrated significant differences between OCD patients and healthy controls.

Conclusions:

• Nonlinear features of EEG microstate sequences offer valuable insights into brain dynamics in OCD.
• These nonlinear features represent potential electrophysiological biomarkers for distinguishing OCD patients.
• The findings support the use of advanced EEG analysis for identifying neurophysiological markers in psychiatric disorders.