WS1.9. Advances in EEG Analysis – Wide-Band EEG, Dense-Array EEG and Quantitative EEG

Abstract

1) Wide-band EEG As the faster frequency extreme of wide-band EEG, high frequency oscillation (HFO) was originally described by means of 1) micro-electrode recording in animal model, and ictal HFO occurred before clinical onset. Recently, 2 more types of HFO were well described; 2) invasive electrodes recording in epilepsy surgery (macro-invasive electrodes such as depth-, subdural ones), and 3) scalp electrodes or MEG in epilepsy patients. The multiple mechanisms of HFO generation are considered such as synchronized post-synaptic potentials with sparse pyramidal cell firing, or principal cell action potentials. As the lower frequency of extreme of baseline shifts in the chronic epilepsy focus on epilepsy patients, ictal DC shifts occurred earlier than or as early as ictal HFO significantly (Imamura et al., 2011; Kanazawa et al., 2015). It was newly endorsed by the pathological abnormality, i.e., decreased Kir4.1, selectively seen in the cortices generating ictal DC shifts (Kobayashi et al., 2017). It could suggest more active role of glia before seizure generation, and thus called as active DC shifts. These both findings could represent activity of both epileptic neurons and astrocytes, respectively, mainly as the local field potentials. 2) Quantitative EEG Once digital EEG era has started since 1989, first generation EEG (1989-1996/2000) could provide us with data acquisition of ordinary sampling rate (i.e., 500 Hz) and band pass filter (i.e., 0.5 Hz - 200 Hz). Data display has been much flexible by means of reformat of montage, filter change, sensitivity change, voltage mapping, power spectrum analysis, and 3D source analysis, and so on. It is user friendly, and regarded as the established tool or method. Second generation EEG (1996/2000-present) has been much advanced and complicated tool such as data acquisition such as wide-band EEG, i.e., 0.016 Hz - 600 Hz. It contains DC shifts, infraslow, HFO. Increased sampling rate, i.e., 2000 Hz could widen the extremely faster activitty. Data display such as wide-band display, i.e., 100 ms/page to 5 min/page, time frequency analysis, DSA (Density Modulated Spectral Array), and Ganger causality further led us to the interactive sophisticated circumstances in the clinical and scientific field. 3) Dense-array EEG The more electrodes are placed on the scalp, the better spatial resolution is expected. Once it also could cover the so-called “south hemispheres” of the head, it could delineate the any activity arising from the basal temporal and mesial temporal areas. However, since scalp EEG is attenuated in amplitude and is blurred with wider distribution by means of cortical bone and different soft tissues, the degree of localization map may not be proportionally precise in spite of involvement of many electrodes. In this regard, whole MEG data may overcome of the 2 important factors.