Changes in EEG-derived entropy are associated with symptom improvement in Major Depressive Disorder following electroconvulsive therapy

Background: Major Depressive Disorder (MDD) is associated with alterations in brain activity that are not fully captured by symptom-based diagnostic approaches. EEG-derived entropy measures have been proposed as potential biomarkers of depressive symptomatology, although findings remain mixed and longitudinal evidence using clinically feasible EEG setups is limited. Methods: Seventeen patients with MDD undergoing modified electroconvulsive therapy (mECT) and seventeen matched healthy controls were included. Resting-state EEG was recorded using a simplified two-electrode frontal montage prior to seizure induction. To maximize data usability, a signal patching procedure was applied to reconstruct segments affected by artifacts while preserving the fundamental properties of the original EEG signals. Entropy was computed across several frequency bands using a permutation entropy calculation with order 3. Changes in entropy between baseline and post-treatment were examined with respect to changes in depressive symptoms. Results: At baseline, patients exhibited higher EEG entropy than controls, particularly in the beta frequency band. Following mECT, patients showed a significant reduction in beta-band entropy. Reductions in left frontal beta-band entropy accounted for variance in symptom improvement. Conclusions: Reductions in frontal beta-band entropy were associated with clinical improvement following mECT and may reflect normalization of cortical activity in MDD. The combination of entropy-based measures with signal patching may enhance the amount of useful information extracted from simplified EEG recordings, supporting their potential utility as biomarkers of treatment response in depression.