Exploring new EEG parameters in electroconvulsive therapy

BackgroundElectroconvulsive therapy (ECT) is an effective treatment against severe depressive episodes, which has been shown to induce volume changes in the hippocampus. The power spectrum of the electroencephalogram (EEG) follows a characteristic power-law relation but its utility as a metric of ECT-induced seizures has not been explored. ObjectiveThis study aims to evaluate a novel metric based on the power spectrum of the EEG recordings from ECT-induced seizures; its association to volume changes in the hippocampus following ECT and improvement in depression rating scores. MethodsDepressed patients treated with ECT underwent brain MRI before- and after treatment and the EEG from each seizure was recorded (N=29). Hippocampal volume changes and EEG parameters were recorded in addition to clinician-rated and self-reported measures of depressive symptoms. The slope of the power-law in the power spectral density of the EEG was calculated. Multivariate linear models relating seizure parameters to volume change or clinical outcome was systematically and successively simplified. The best models were selected according to Akaikes information criterion. ResultsThe slope of the power-law was steeper in the right than the left hemisphere (p < 0.001). EEG measures were included in the best models of volume change for both hippocampi as well as in the models explaining clinical outcome (p = 0.014, p = 0.004). ConclusionsA novel EEG measures was explored and contributed in models explaining the variation in volume change in the hippocampus and in clinical outcome following ECT.