An adaptive algorithm for real-time electrode calibration

Continuous brain monitoring based on EEG recorded from surface electrodes is believed to have potentials in wearable medical devices. In such devices capacitive electrodes are attractive compared to conventional electrodes because there is no need for skin preparation and conductive gels, and because of diminished motion artifacts. However, there are technical challenges connected to the practical application of capacitive electrodes. The electrode capacitance, which has significant impact on the signal measured, will vary between channels and will be time varying. Therefore calibration of the electrode array is an important preprocessing step before the signal processing. This paper proposes an algorithm for blindly estimating the parameters of the analog signal acquisition paths, including the capacitances of the electrodes. The algorithm continuously estimates the parameters, based on the measured EEG signals, and compensates for variations in the analog signal paths. Simulations show that the algorithm can estimate the parameters, and track changes of the electrodes capacitance in real-time.