A comparison of the power spectral density of scalp EEG and subjacent electrocorticograms.

Objective Our study investigated the effects of the intact skull on background EEG rhythms recorded simultaneously by subdural electrocorticography (ECoG) and scalp EEG. Methods We performed a retrospective analysis of twenty patients undergoing intracranial EEG monitoring. EEG and ECoG were recorded simultaneously from the central and occipital scalp and subjacent subdural electrodes removed (median 46 mm, interquartile 27–65) from the craniotomy. The power spectral density (PSD) of artifact-free EEG and ECoG segments and ratio of the scalp EEG to subjacent ECoG PSD was calculated. Results Overall both ECoG and scalp EEG power decreased by over three orders of magnitude from delta to gamma frequency band with an empirical inverse power relationship. The ratio of scalp EEG to ECoG PSD decreased across the delta and theta frequency bands, remained the same across the alpha, beta and low gamma bands, but increased at the higher frequency bands. Conclusions EEG PSD mirrored changes in ECoG PSD across the frequency bands. As ECoG power continued to decrease above 42 Hz, extracranial voltage sources contributed to a greater fraction of scalp EEG power. Significance Monitoring the gamma frequency band using scalp EEG was limited by low power on ECoG and masking by extracranial voltage sources. [ABSTRACT FROM AUTHOR]