Classification of obsessive compulsive disorder by EEG complexity and hemispheric dependency measurements

Area of Science:

• Neuroscience
• Computational Psychiatry
• Biomedical Engineering

Background:

• Obsessive-compulsive disorder (OCD) is linked to orbitofronto-striatal dysfunction.
• Previous research suggests OCD may be a subgroup of anxiety disorders with altered brain complexity.

Purpose of the Study:

• To investigate single-channel electroencephalography (EEG) complexity and interhemispheric dependency for OCD classification.
• To identify specific EEG features differentiating OCD patients from controls.

Main Methods:

• Utilized 19-channel eyes-closed EEG recordings.
• Applied permutation entropy (PermEn), approximate entropy, and sample entropy for EEG complexity.
• Examined mean coherence and mutual information for interhemispheric dependency.
• Employed support vector machine classifiers with 2-fold cross-validation on 2-second EEG segments.

Main Results:

• Permutation entropy (PermEn) at prefrontal regions yielded the highest classification accuracy (85 ±5.2%).
• Other entropy measures also showed significant differences between OCD patients and controls in prefrontal regions, except for coherence.
• OCD patients exhibited lower EEG complexity in prefrontal and right fronto-temporal areas.

Conclusions:

• Functional abnormalities in pre-frontal regions contribute to OCD.
• Reduced EEG complexity in specific brain areas is a characteristic of OCD patients.
• Findings align with neuroimaging studies classifying OCD within anxiety disorders with decreased neural complexity.