Nonlinear stimulus artifact cancellation in somatosensory evoked potential studies

Somatosensory evoked potentials (SEPs) are clinically important bioelectric signals. The surface measurements of these potentials are often contaminated by the stimulus artifact which, depending on the stimulator and measurement system characteristics, may mask the initial portion of the SEP. Here, a novel approach is undertaken to diminish this artifact. A nonlinear adaptive filter (NAF) based on the truncated second order Volterra series is employed to iteratively minimize the artifact. The effectiveness of the NAF is demonstrated qualitatively using the experimental data collected from the wrist region and the spinal cord region.