Models for the Quantitative Prediction of Therapeutic Responses Based on the Baseline EEG Parameters in Depressive Patients

Abstract

To determine the possibility for individualized quantitative prediction of therapeutic responses in depressive patients from the baseline EEG parameters, we conducted a correlational analysis of relationships between the baseline EEG spectral power values (80 variations in total) recorded in 42 depressive patients before the start of the therapeutic course and clinical quantitative assessments of the post-treatment mental conditions of these patients. Based on these data, regression models were built for individualized quantitative prediction of therapeutic response, including no more than three pre-treatment EEG parameters and describing up to 75% of the variance in clinical post-treatment assessment values. On the one hand, our results confirm the possibility of designing fairly accurate mathematical models for the individualized quantitative prediction of therapeutic responses in depressive patients by a small number of baseline neurophysiological parameters. On the other hand, the employed mathematical approaches make it possible to clarify the neurophysiological mechanisms underlying depressive disorders.