Basic linear filters in extracting of auditory evoked potentials

Abstract

The aim of this study is to assess the performance of additivity-based linear filtering techniques into two groups in extracting of auditory Evoked Potentials (EPs) from a relatively small number of sweeps. We named these groups as: Group A (The Wiener Filtering (WF) and coherence weighted WF (CWWF)) of orthogonal projections) and Group B (standard adaptive algorithms of Least Mean Square (LMS), Recursive Least Square (RLS), and one-step Kalman filtering (KF)). All methods are compared to the traditional ensemble averaging (EA) in simulations, pseudo-simulations and experimental studies based on the signal-to-noise-ratio (SNR) enhancement. We observed that the KF is the best methods among them. The filtering of the projections instead of the raw data improves the performance of filtering operations in both cases of the LMS and WE The CWWF works better than the conventional WF when it is applied to the projections as well. In conclusion, most of the linear filters show definitely better performance compared to EA. The KF effectively reduce the experimental time (to one-fourth of that required by EA). The projection method so called Subspace Method (SM) in the current study is a useful pre-filter to significantly reduce the noise on the raw data. The use of the SM is revealed in auditory EP estimation. The SM improves the performance of different algorithms.