Analysis of the Generation of Auditory Steady-State Cortical Evoked Responses in Guinea Pigs

Recent research shows that human auditory steady-state responses (ASSRs) develop a resonance at 40 Hz and the dramatic amplitude increase of the Pb component of the middle latency response (MLR) accounts for the high amplitude of the ASSR at 40 Hz. The first part of this study aimed to investigate the ASSR resonance characteristics as a function of rate in guinea pigs. A study of the grand average of the peak-to-peak and fundamental frequency amplitudes does indeed show a resonance around 40 Hz in guinea pigs. Unlike human ASSRs, this resonance is very broad (26-52 Hz) and flat. The centrally recorded ASSRs are smaller and tend to have resonances at higher rates compared to temporal signals. The second part of the analysis investigated whether the superposition of transient responses can predict the acquired ASSRs at each corresponding rate. This superposition theory is one of two competing theories on the origin of the ASSRs, with the other centering on the induced phase synchronization of brain waves. In order to test the first theory, transient responses were used to create synthetic ASSRs, which were then compared to the acquired ASSRs via correlation coefficient and phasor analysis. For the 40 Hz ASSR, both temporal and central electrode synthesized ASSRs show a correlation coefficient above 0.80. In the comparison at 20 Hz, the correlation coefficient is very high (about 0.9) in the temporal electrode, yet significantly lower (about 0.7) for the central electrode. Furthermore, at 80 Hz, the correlation coefficient is significantly lower in both temporal and central electrodes (about 0.7). At all rates, the correlation coefficients are highest with low jitter sequences. Finally, phasor analysis was also used to test the superposition theory of the generation of the acquired ASSRs at 20, 40, and 80 Hz. Overall, in the temporal recordings at 40 Hz, the superposition of the MLR responses accurately predicted the acquired 40 Hz ASSR as demonstrated by both magnitude and phase analysis. The recordings made in the central electrode only predicted the acquired ASSR in its phases, with significant differences found in magnitude at its main harmonics. Similarly, at 20 and 80 Hz in both temporal and central electrodes, the synthetic ASSRs did not appear to fully predict the acquired ASSRs. Although the phases were successfully predicted, large magnitude variations were observed. As shown by mean prediction error plots, the acquired ASSRs are best predicted by low jitter sequences, followed by low-medium and medium jitter sequences