Cochlear implant artifact reduction on one channel Mismatch Negativity recordings based on Ensemble Empirical Mode Decomposition and Independent Component Analysis

Artifact generated by cochlear implants has been a problem for being able to register Mismatch Negativity (MMN) response. There are methods for reducing the artifact using multiple channels from the EEG but in this paper are presented the first results of a method using only the channel with the artifact using Ensemble Empirical Mode Decomposition (EEMD) and Independent Component Analysis (ICA). The first results showed that it was possible to get the MMN registers from the group of normal recordings and partially with the group of recordings from patients with cochlear implant. It is possible to suggest that EEMD in conjunction with ICA can be used for studies searching MMN.