Abstract
Purpose :
To investigate the Visual Evoked Potentials (VEPs) adaptation at high rates.
Methods :
Monocular Pattern Electroretinograms (PERGs) and Visual Evoked Potentials (VEPs) were simultaneously acquired at stimulation frequencies: 2, 10, 15, 20 and 25 reversals per second (rps) in 7 normal subjects (18-35yo). Two stimulation paradigms were used: constant rate and slightly-jittered sequences at same rates. Constant rates generated steady state responses (SSR) that were analyzed in frequency domain. Jittered rates generated quasi-SSR that were deconvolved to retrieve transient responses. The spectra of the raw recordings was analyzed to assess the influence of the stimulation on the background neural activity. In a second experiment, 32s long recordings consisting of 8s at 2rps followed by 8s at 25rps and return to 8s at 2rps were conducted to study the temporal dynamics of the neural activity. Time-frequency analysis was performed with the Morlet wavelet transform.
Results :
The maximum amplitudes for steady state PERG (SS-PERG) and steady-state VEPs (SS-VEP) were obtained at 15 and 10rps, respectively. Spectra of SSR raw recordings showed enhanced spectral peaks at stimulation frequencies and its harmonics. High rate steady state VEPs (>2rps), on the other hand, produced enhanced background Electroencephalogram (EEG) power at alpha-band (~10Hz) compared to low rate (2rps) responses; this effect was not observed on the PERG spectra. The time-frequency analysis on the multi-rate recordings verified the increase of the EEG alpha power (~25%), synchronized with the onset of the high-rate stimulation and a return to 2rps baseline.
Transient VEPs obtained by deconvolution at 15rps and higher were of much higher amplitudes compared to corresponding SS-VEPs. Synthetic SS-VEPs computed using deconvolved transient responses successfully predicted the acquired steady-state responses.
Conclusions :
Rate characteristics of transient and SS-PERGs were consistent with previous studies. Deconvolved transient VEPs, however, revealed no amplitude reduction at high rates as expected from analysis of SS-VEPs. Superposition models for the generation of evoked responses predict the spectra enhancement at stimulation rates and harmonics well but not the rate-dependent alpha power enhancement. This increase may help to explain the amplitude of the high-rate transient VEP. Results are relevant for better understanding of the VEP generators.
This is an abstract that was submitted for the 2016 ARVO Annual Meeting, held in Seattle, Wash., May 1-5, 2016.