Abstracts

Rationale: Tripolar concentric ring electrode (TCRE) electroencephalography (tEEG) was first introduced by Besio et al. [1]. Two bipolar signals from the TCRE are combined to form tEEG signals derived with 16[(Middle - Disc) - (Outer - Disc)] the Laplacian algorithm first described in [1]. Where Disc is the central disc, Middle is the middle ring, and Outer is the outer ring of the TCRE. We have shown that compared with disc signals, tEEG has nearly 4-fold (374%) the signal to noise ratio (SNR) and less than one-tenth (8.27%) the mutual information [1, 2]. TCRE has also strong attenuation of common-mode artifacts [1]. Because of these advantages we sought to detect high-frequency oscillations (HFOs) during peri-seizure periods from TCRE placed on the scalp surface. Methods: We recorded electrographic activity from three patients with epilepsy by placing a set of TCREs in the prime 10-10 system locations in addition to the standard 10-20 system electrode locations used in the hospital. The conventional disk EEG (dEEG) and tEEG were both recorded at 200 S/s (hardware filters: dEEG 1- 70 Hz, tEEG 1-100 Hz respectively). We selected approx. 30 min segments, centered on the onset of the generalized tonic seizure, and generated spectrograms to follow the temporal dynamic of pre-seizure gamma-band HFOs. Results: Fig. 1 shows representative dEEG and tEEG signals. Panel B shows bipolar dEEG (Fp2-F4). Panel D shows tEEG recorded from location Fp2' of the 10-10 system directly behind the Fp2 conventional electrode. Panels A and C are the corresponding time-frequency spectrograms calculated for 2 sec sliding windows. In the magnified traces E and F an additional 0.7 Hz high-pass filter was applied to remove a very low frequency drift component. In two patients with generalized seizures we found specific locations where the tEEG recorded HFO activity in the gamma band (~70 Hz)approximately five to ten minutes before the seizure. The electrode locations exhibiting HFOs were within the seizure onset zone determined by the neurologist (IEMJ). The red ellipse in Panel C highlights the HFO in the TCRE at location Fp2'. HFOs were not found in dEEG even in cases when HFOs recorded by tEEG were at frequencies <70 Hz (the upper filter setting for dEEG). In the third patient who did not have a seizure,the spectrogram revealed periods of frequent interictal spikes when gamma-band HFOs were also present. Conclusions: For the first time we report records of human ictal activity captured with tEEG. These preliminary data indicate that tEEG is capable of detecting HFOs from scalp, which are not usually seen in conventional EEG records.Thus, tEEG may provide additional tools to follow peri-seizure dynamic with greater precision especially in the high frequency range.