A comparison of methods for analysis of temporal frequency tuning in visual responses

INTRODUCTION: There are numerous methods to study neuronal processing of information about temporal frequency content of visual stimuli. The two most fundamental methods are 1) direct measurement of response amplitude, e.g. an amplitude of averaged visual evoked potential, and 2) assessment of response magnitude after transformation of electrophysiological signal from time to frequency domain. AIM(S): The aim of this study was to find an appropriate analysis method to characterize cortical responses to visual stimuli of various temporal frequencies. METHOD(S): Visual responses were recorded from both primary visual cortices, contra- and ipsilateral to the stimulated eye, using multichannel linear electrode arrays during electrophysiology experiments performed on anesthetized rats. As a visual stimulus we used 2-ms-long LED flashes delivered at two frequencies: 1 and 7 Hz. RESULTS: We found that for frequency of 1 Hz it is difficult to draw conclusions based on power spectrum alone. For frequency of 7 Hz the assessment of evoked potential in time domain was highly inaccurate. CONCLUSIONS: For 1 Hz the estimation of the visual evoked potential amplitude by direct measurement should be also performed. For 7 Hz the analysis should be performed after transformation of the signal from the time to frequency domain. Our results also indicate the advantages of the Welch method in comparison to the periodogram to analyze signals in the frequency domain. FINANCIAL SUPPORT: Supported by the Polish National Science Center grant Symfonia 1 (2013/08/W/NZ4/00691).