Reducing the noise level of EEG signal acquisition through reconfiguration of dry contact electrodes

Dry EEG electrodes enhance the convenience and wearability of brain computer interface (BCI) systems, but the noise induced due to the skin electrode interface reduces the signal quality compared to that of wet electrodes. In this work, we analyzed a popular dry electrode design that incorporates several `finger´ shaped contacts and show how the noise level can be reduced by estimating the contact quality of each finger and reconfiguring the electrode accordingly. We designed custom electrodes for this analysis and showed the differences in skin interface noise among fingers of the same electrode. Finally, we showed how eliminating the signals from one of more noisy fingers can lead to better correlation with an ideal wet electrode and reduce the amount of noise by 1.6μV on average, which constitutes 11% of the EEG signal.