Ictal EEG Non-linear And High Order Spectral Analysis Methods In Electroconvulsive Therapy And Its Clinical Utility

Area of Science:

• Neuroscience
• Computational Psychiatry

Background:

• Electroencephalogram (EEG) analysis during electroconvulsive therapy (ECT) can predict clinical outcomes.
• EEG signals are complex, non-linear, and non-stationary, requiring advanced analytical methods.
• Non-linear and higher-order spectrum analyses offer better characterization of EEG signals compared to traditional methods.

Purpose of the Study:

• To investigate the predictive value of non-linear and higher-order spectrum analyses of ictal EEG for clinical outcomes in schizophrenia patients undergoing ECT.
• To determine if specific EEG measures recorded early in the ECT course can forecast treatment response.

Main Methods:

• Schizophrenia patients undergoing ECT had their EEG recorded during seizures.
• Non-linear measures including Approximate Entropy (ApEn), Sample Entropy (SamEn), Hurst exponent (H), correlation dimension (CD), and Largest Lyapunov exponent (LLE) were computed.
• Higher-order spectrum analysis, specifically Bispectrum entropy (HOS. En), was also calculated from early ECT EEG sessions.

Main Results:

• Bispectrum entropy (HOS. En) significantly predicted clinical outcome at two weeks (r = -0.434, p = 0.027 and r = -0.414, p = 0.036).
• Other non-linear measures like ApEn, SamEn, H, CD, and LLE did not show a significant correlation with clinical outcome.

Conclusions:

• Higher bispectrum entropy of ictal EEG during early ECT sessions is a significant predictor of better clinical outcomes in schizophrenia patients.
• This finding highlights the potential of advanced EEG analysis techniques for personalized ECT treatment strategies.