Abstracts

Rationale: The purpose of our study was to investigate the effects of the intact skull on background EEG rhythms of patients undergoing long term combined scalp and intracranial EEG monitoring. Multiple studies reported an inverse relationship between frequency and power recorded using subdural electrodes (ECoG); the power spectral density (PSD) decreased by four orders of magnitude from 1 Hz to 125 Hz. Overall scalp EEG PSD decreased in parallel to the subdural PSD for the Berger frequency bands (1 Hz to 25 Hz). We previously reported that the ratio of scalp to subdural PSD measurements for paired electrodes ranged from 13% in the delta band to 6% in the alpha and beta bands. We now report on paired subdural and scalp PSD measurements in the gamma and omega bands (25 to 125 Hz).Methods: Twenty patients (age 18-55) from the Yale University Epilepsy Surgery Program undergoing intracranial EEG monitoring for surgical evaluation were recruited. Scalp EEG recording were made using Grass platinum needle electrodes located over the C3, C4, O1 and O2 sites. Simultaneous recordings were obtained from subjacent subdural strip electrodes (4mm diameter platinum disks with 2.3mm exposed surface diameter). Offline analysis was performed with custom software written in a mixture of high level languages and MATLAB. Scalp and intracranial EEG epochs, 1 hour in duration, at least 6 hours removed from a seizure, 3 or 4 days after electrode implantation surgery, during wakefulness between 9-10 AM were selected for analysis. Scalp and subdural recordings were examined for artifacts and segmented at a 1-s resolution. The power of artifact-free EEG segments were obtained for each electrode contact studied and averaged over the epoch. Subdural and scalp PSD of paired electrodes were calculated for the 0.5 and 128 Hz frequency range.Results: As previously reported subdural PSD continued to decrease steeply as frequency increased. A single power function (PSD = frequency-3.1) described the relationship from the theta band through the omega band. Similarly, the scalp PSD may be described by a homologous function over the alpha, beta and low frequency gamma bands. Above 80 Hz the scalp PSD did not decrease in parallel with the subdural PSD and approached a linear asymptote – a nearly constant value. By 125 Hz the interquartile (95% confidence interval) of the subdural and scalp PSD medians began to overlap.Conclusions: There was a greater transfer of power from cortical source signals to scalp EEG signals in the delta through beta frequency bands than in higher frequency bands. At frequencies above 80 Hz, the relationship between the power of scalp and subdural EEGs was dominated by extracranial and artifactual signals recorded by the scalp EEG. The scalp EEG power did not reflect accurately the subdural ECoG signals generated by the human cortex.