EEG-induced Effective Connectivity Analysis in Major Depressive Disorder | JoVE Visualize

Area of Science:

Neuroscience
Psychiatry
Biomedical Engineering

Background:

Depression poses a global mental health challenge, with current diagnostic methods limited by subjectivity and inaccuracies.
Existing research often focuses on functional connectivity, overlooking causal interactions between brain regions.
There is a critical need for objective neural biomarkers to improve the diagnosis and treatment of Major Depressive Disorder (MDD).

Purpose of the Study:

To investigate effective connectivity (EC) patterns in resting-state Electroencephalography (rsEEG) signals of individuals with MDD compared to healthy controls (HC).
To apply the Frequency-Domain Convergent Cross Mapping (FD-CCM) technique to capture EC in the frequency domain, addressing limitations of previous methods.
To evaluate the potential of FD-CCM-derived features as diagnostic biomarkers for MDD.

Main Methods:

Utilized resting-state Electroencephalography (rsEEG) data from individuals diagnosed with Major Depressive Disorder (MDD) and healthy controls (HC).
Applied the Frequency-Domain Convergent Cross Mapping (FD-CCM) technique, a nonlinear, model-free method, to analyze effective connectivity patterns.
Compared FD-CCM features against classical Cross Recurrence Quantification Analysis (CCM) and employed Artificial Neural Network (ANN) classifiers for diagnostic accuracy assessment.

Main Results:

MDD subjects exhibited significantly reduced EC across frontal, parietal, temporal, and occipital brain regions compared to HC participants.
Diminished frontal connectivity and altered power density in delta and alpha frequency bands were observed in the MDD group.
FD-CCM features achieved a superior classification accuracy of 92.32% using an ANN classifier, outperforming classical CCM methods.

Conclusions:

Altered effective connectivity patterns serve as significant neural biomarkers for Major Depressive Disorder (MDD).
These connectivity deficits are associated with impairments in cognitive processing, emotional regulation, and sensory integration in MDD.
The Frequency-Domain Convergent Cross Mapping (FD-CCM) approach demonstrates high potential for clinical applications in diagnosing mental health conditions like MDD.