Cerebral potentials evoked by Co, laser stimulation were recorded in 22 normal volunteers. Substantial variability in latency. morphology and amplitude of the responses in individual trials was noted, and in some trials no response appeared to be present. Averaging after latency correction and removal of traces without clear responses was compared with conventional time-locked averaging. We found a greater amplitude and more consistent wave form in the former case, and better preservation of amplitude with increasing age and under conditions of impaired attention. Cross-correlation techniques and Minimum Mean Square Error (MMSE) filtering did not assist identification of responses. We conclude that with laser-evoked cerebral potentials visual identification of components is a simple and reliable way of obtaining an improved estimate of the average evoked potential.
