Interaction between steady-state visually evoked potentials at nearby flicker frequencies

Siddhesh Salelkar,Supratim Ray

doi:10.1038/s41598-020-62180-y

Abstract

Steady-state visually evoked potential (SSVEP) studies routinely employ simultaneous presentation of two temporally modulated stimuli, with SSVEP amplitude modulations serving to index top-down cognitive processes. However, the nature of SSVEP amplitude modulations as a function of competing temporal frequency (TF) has not been systematically studied, especially in relation to the normalization framework which has been extensively used to explain visual responses to multiple stimuli. We recorded spikes and local field potential (LFP) from the primary visual cortex (V1) as well as EEG from two awake macaque monkeys while they passively fixated plaid stimuli with components counterphasing at different TFs. We observed asymmetric SSVEP response suppression by competing TFs (greater suppression for lower TFs), which further depended on the relative orientations of plaid components. A tuned normalization model, adapted to SSVEP responses, provided a good account of the suppression. Our results provide new insights into processing of temporally modulated visual stimuli.

Highlights

Steady-state visually evoked potential (SSVEP) studies routinely employ simultaneous presentation of two temporally modulated stimuli, with SSVEP amplitude modulations serving to index top-down cognitive processes
As expected for counterphase stimulation, there was a prominent SSVEP response at double the stimulation frequency (2F), which was visible when averaged across the population of spiking electrodes (Fig. 1c; electrodes were selected from each session using predetermined cutoffs before being pooled together, see Methods; n = 32; Monkey 1: n = 20; Monkey 2: n = 12) and local field potential (LFP) electrodes (Fig. 1d; electrodes having consistent LFP responses and stable estimates of receptive field across days; n = 650; Monkey 1: n = 65 electrodes × 7 sets across 6 orientations; Monkey 2: n = 39 electrodes × 5 sets across 4 orientations) at full contrast
Since SSVEP response amplitudes in spiking and LFP were found to peak around the 8–16 Hz range (Fig. 1), we presented one of the gratings at either 8 Hz or 16 Hz (“target” frequency, fTARGET, separate sessions for each), while the temporal frequency (TF) of the other grating (“mask” frequency, fMASK) varied in a close range around fTARGET between lower delta (2 Hz) to high beta (30 Hz), depending on the fTARGET used in that session

Summary

Introduction

Steady-state visually evoked potential (SSVEP) studies routinely employ simultaneous presentation of two temporally modulated stimuli, with SSVEP amplitude modulations serving to index top-down cognitive processes. SSVEP has often been used in masking paradigms, in which human observers detect a temporally modulated target grating when it is superimposed by another temporally modulated mask grating presented at different contrasts and orientations relative to the target[13,14,15,16] These studies have shown that the target SSVEP is reliably suppressed in the presence of the mask, depending on its contrast and relative orientation. Despite the widespread use of SSVEP, its precise mechanisms remain unclear, the dominant view holds that the SSVEP results from the synchronized activity of spatially homogeneous cortical neurons in an open-field arrangement[31] This is a simplified assumption, since the surface EEG can relate to the underlying neural activity in more complex ways[32,33,34,35]. Attempts to probe the nature of TF normalization at multiple scales of neural recording, could potentially shed light on the mechanisms of SSVEP generation as well

Objectives

Methods

Results

Conclusion