The effects of automatic artifact rejection on evoked potential recordings

An analysis of the effects of automatic artifact rejection on the evoked potential recordings is made using both a theoretical analysis and computer simulations. It is demonstrated that when the noise at each measurement point is independently distributed with maximum probability near zero amplitude, the bias and changes in variance induced by the automatic artifact rejection are small unless more than 80% of traces are rejected. When high-amplitude noise transients occur, there can be significant bias in the evoked potential when the artifact rejection threshold is set near the amplitude of the noise transients. Even more substantial distortions in the evoked potential occur when the noise is not independently distributed at each measurement point. It is important to recognize the possibility of these effects during clinical evoked potential recording.