Engineering Aspects of the Quantified Amplitude-Integrated Electroencephalogram in Neonatal Cerebral Monitoring

Abstract

Amplitude-integrated electroencephalography (aEEG) is used for determination of severity of hypoxic-ischemic encephalopathy and eligibility for hypothermia therapy in infants. We evaluate susceptibility of the aEEG to machine differences, impedance mismatch, electrode placement, and interelectrode distance in preparation for further clinical studies. Frequency-response curves were obtained from manufacturers of aEEG systems. The mean of the quantified aEEG and the median lower border of amplitude were derived from digital EEG recordings. Impedance effects were evaluated by inserting resistors ranging from 1 kOmega to1 MOmega into inputs G1/G2 while recording from a looped-signal generator. Interelectrode distance effects were evaluated from simultaneous recordings from electrodes placed at 5% to 30% of the preauricular notch distance from a healthy human subject. Electrode location effects were evaluated by comparing Q-aEEG from C3-C4 versus P3-P4 from 10 term infants. Frequency-response curves differed up to 20 dB for particular frequencies. Impedance mismatch distorted EEG and aEEG, but did not significantly change Q-aEEG. Increasing interelectrode distances significantly increased Q-aEEG. Central Q-aEEG values were significantly higher than parietal. Comparisons of aEEG must carefully account for differences in machines, interelectrode distances, and electrode locations. Q-aEEG may offer a rigorous method of evaluating hypoxic-ischemic encephalopathy.