An algorithm to remove artifacts from EEG based on adaptive FL-BPNN filter

Electrooculogram and Electromyography are the most common interference in electroencephalogram (EEG). In this paper, a new adaptive FL-BPNN-based filter is proposed to cancel the two most serious contaminants. In the proposed filter, the functional link neural network is applied to the consequent part of the fuzzy rules, and thus the consequent part of the proposed model is a nonlinear combination of input variables that increases its universal nonlinear approximation ability considerably. This paper presents an adaptive noise cancellation algorithm to remove artifacts in EEG based on adaptive FL-BPNN filter. The algorithm takes the least squares estimation to adjust antecedent parameters, and takes the hybrid algorithm of back propagation algorithm to adjust consequent parameters. The experimental results indicate that the performance of the proposed filter is better compared with the ANFIS filter. Thus the algorithm is effective for removing EOG and EMG in EEG.