Estimation of Visual Evoked Potentials using a Signal Subspace Approach

Extraction of visual evoked potentials (VEPs) from the humanbrain is generally very difficult due to its poor signal-to-noise ratio (SNR) property. A signal subspace technique is presentedto estimate VEPs hidden inside highly colored electroencephalogram (EEG) noise. This method is borrowed andmodified from signal subspace techniques originally used forenhancing speech corrupted by colored noise. The signalsubspace is estimated by applying eigenvalue decomposition on the approximated signal covariance matrix. The signal subspacebased algorithm is able to satisfactorily extract the P100, P200and P300 peak latencies from artificially generated noisy VEPs.The simulation results show that the estimator maintains anaverage success rate of 87 % with an average percentage error of less than 9 %, when subjected to SNR from 0 dB to -10 dB.