In linear systems analysis, a system is often studied by determining its response to impulsive stimulation. Alternatively, the system's response to continuous noise may be measured; cross-correlation of input and output then yields a function identical to the impulse response. This approach has been extended to auditory evoked responses, using vertex potentials as the system output. Although the auditory system, viewed as a brain stem evoked response generator, is highly nonlinear, replicable cross-correlograms were obtained in a series of guinea pigs. Stimulus artifact was ruled out by comparison of these responses with autocorrelograms of the noise stimulus. Responses decreased in amplitude with decreasing stimulus intensity. Experiments involving brain stem lesions and anoxia suggested that the short-latency cross-correlogram reflect cochlear microphonic, auditory nerve, and brain stem contributions. Study of bioelectric potentials by cross-correlation with noise input allows data acquisition continuously, without the limitations in stimulus repetition rate and dynamic range imposed by impulse (eg, click) analysis. This approach may prove useful in both experimental and clinical settings.
ASJC Scopus subject areas