Date of Completion

January 1987


Engineering, Biomedical




The motivation for this research was to develop an evoked potential methodology for non-invasively monitoring the auditory and visual sensory channel engagement and interaction (A x V) in humans. Sensory interaction was concluded whenever response variations of one sensory channel could be attributed to parameter changes of an additional stimulus, simultaneously presented to another sensory channel. This study adapted the phase-lock technique in characterizing the stimulus parameter-space of the A x V process.^ The stimuli consisted of sinusoidally modulated, 100% amplitude-modulated tone and spot-light signals. The independent variables were the stimulus modality (audio, visual or both), Modulation Frequency (MF, 5-61 Hz), intensity (24 dB range) and subject attentiveness. Continuous Evoked Potentials (CEPs) were recorded from Cz-A1 (auditory) and Oz-A1(visual) sites. The CEP dependent variables were the magnitude and phase of the evoked potential component phase-locked to the stimulus fundamental or 2nd harmonic MF. The CEPs were characterized in terms of their magnitude, phase and latency Modulation Transfer Functions, background EEG, repeatability, linearity, and stimulus intensity and subject attentiveness dependencies. Finally, the A x V process was assessed on a selected stimulus parameter-space.^ Most of the results were derived from two MF regions, "Beta" (16-25 Hz) and "Theta" (5-6 Hz). Delays and latencies of Beta and Theta CEPs were estimated from linear models fitted to the magnitude data; 30-40 msec and 60-70 msec (Beta), and 50-60 msec and 200-240 msec (Theta), respectively. Responses elicited from these MF regions were consistent, exclusively evoked, repeatable and easily detectable. In comparison to the auditory CEPs, the visual CEPs were associated with higher signal/noise magnitude ratio, higher inter-subject variability and lower intra-subject variability.^ Genuine A x V effects of 5.5 dB magnitude-gain and 40 degree phase-shift were found. Magnitude inhibition and phase lag were generally associated with, and induced by, the short-latency Beta MF region. Magnitude data was primarily affected by "Internal" parameters (intra-modality stimulus intensity or MF) and phase data by "External" parameters (cross-modality stimulus intensity). Attention allocation did not play a major role. ^