Graph-based analysis of brain connectivity in schizophrenia

Area of Science:

• Neuroscience
• Psychiatry
• Biomedical Engineering

Background:

• Schizophrenia is a complex psychiatric disorder associated with disruptions in brain function.
• Electroencephalography (EEG) is a valuable tool for investigating brain activity and connectivity.

Purpose of the Study:

• To evaluate brain connectivity differences between schizophrenia patients and healthy controls using EEG.
• To compare the effectiveness of different EEG data processing techniques in revealing these differences.

Main Methods:

• EEG data from 14 schizophrenia patients and 14 healthy controls were analyzed.
• Connectivity measures including Phase-Locking Value (PLV), Phase-Lag Index (PLI), and Directed Transfer Function (DTF) were calculated.
• Graph theory indices were applied to analyze connectivity matrices after Current Source Density (CSD) transformation and Reference Electrode Standardization Technique (REST) re-referencing.

Main Results:

• Both CSD and REST preprocessing methods reduced volume conduction effects.
• The alpha frequency band showed the most significant group differences in connectivity.
• Schizophrenia patients displayed reduced phase synchronization, lower clustering coefficients, and shorter characteristic path lengths.
• Altered directional flow from the occipital region and fronto-posterior asymmetry were observed in patients.

Conclusions:

• Preprocessing techniques like CSD and REST are crucial for accurate EEG connectivity analysis.
• Graph-based analysis of different frequency bands can identify disconnectivity patterns in schizophrenia.
• These findings highlight potential biomarkers for schizophrenia using EEG-derived brain network analysis.