Segmented matched filtering of single event related evoked potentials

Abstract
A fast segmentation-based Matched Filtering (MF) technique of single trial Evoked Potentials (EP's) is presented. MF improves the Signal-to-Noise Ratio of single EP's, reducing the number of repetitions necessary to obtain high quality signals by an order of magnitude. A computer simulation and analysis of experimental data of Movement Related Potentials and cognitive Event Related Potentials demonstrate the superior capabilities of MF compared to traditional Ensemble Averaging.