Quantitative electro-oculography and electroencephalography as indices of alertness

Even though both electro-oculography (EOG) and EEG have been widely used for assessing alertness, the relationship between these two measures has not yet been clarified at various alertness levels. We estimated the frequencies of eye movements faster than 15 degrees/s (NoEM15) and EEG power at every 0.5 Hz step frequency point, quantitatively as well as continuously from alert to very light sleep (Stage 1) in 14 healthy adults. We devised a new EOG derivation for a computerized analysis, because conventional EOG recordings suffer from EEG contamination and have been analyzed manually. EOG electrodes were attached to the left orbitale in this study. The eye movements were detected as peaks in differentiated EOG signals. Significant correlations were found between NoEM15 and EEG powers at several frequency points, mainly in the alpha and beta bands. Though fluctuating EEG components in close correspondence to the frequency of the eye movements varied depending on the individual, EEG power at the peak frequency of the awake state was most closely associated with the eye movements. A broad band power that centers at the peak frequency may be a better measure for alertness assessment than powers in fixed bands.