Investigate the utility of quantitative intracranial EEG for detection of seizure onset and localization compared to standard intracranial tracing analysis.

We randomly selected 7 patients and analyzed their intracranial EEG for a total of 25 focal onset seizures. Seizures were transformed to frequency domain and displayed as FFT power spectrogram. An independent board-certified epileptologist analyzed the colored spectrogram and identified 1. Electrodes involved at seizure onset, 2. Electrode with most prominent change at onset and 3. Time of onset. The epileptologist was blinded from raw EEG tracings and text documentation. Comparison was made via T-test to the electrodes and time at ictal onset reported in the epilepsy monitoring unit (EMU).

Seizure onset time interpreted by qEEG and EEG was averaged to +/- 0.6s of variability, which compared by paired T-test showed no statistical difference for each patient (p=0.56). The earliest prominent electrodes at onset were in agreement in 88% of seizures, and all electrodes with prominent change at onset was in agreement in 100% with reported seizure onset. The number of channel groups with prominent change at onset was significantly larger with qEEG compared to EMU (total 157 vs. 59, p<0.0001).

Reported seizure onset time was not statistically different between qEEG and EEG, and there was high agreement regarding prominent channels of involvement. Too, number of identified channels was statistically larger in the qEEG sample, suggesting higher sensitivity in detection. These findings support qEEG FFT analysis as a useful tool in intracranial EEG interpretation.