Identification of the neural sources of the pattern-reversal VEP

Abstract

This study aimed to characterize the neural generators of the early components of the visual-evoked potential (VEP) to pattern-reversal gratings. Multichannel scalp recordings of VEPs and dipole modeling techniques were combined with functional magnetic resonance imaging (fMRI) and retinotopic mapping in order to estimate the locations of the cortical sources giving rise to VEP components in the first 200 ms poststimulus. Dipole locations were seeded to visual cortical areas in which fMRI activations were elicited by the same stimuli. The results provide strong evidence that the first major component of the VEP elicited by a pattern-reversal stimulus (N75/P85) arises from surface-negative activity in the primary visual cortex (area V1). Subsequent waveform components could be accounted for by dipoles that were in close proximity to fMRI activations in the following cortical areas: P95 (area MT/V5), P125/N135 (area V1), N150 (transverse parietal sulcus, TPS), N160 (ventral occipital areas VP, V4v, and V4/V8), and N180 (dorsal occipital areas V3A/V7). These results provide a detailed spatiotemporal profile of the cortical origins of the pattern-reversal VEP, which should enhance its utility in both clinical and basic studies of visual-perceptual processing.

Introduction

Studies of transient visual-evoked potentials (VEPs) in humans have used stimulus patterns presented briefly in on–off mode (pattern onset) as well as counterphase shifts of gratings or checkerboards (pattern reversal). Pattern-onset VEPs are characterized by an initial component at 70–90 ms after stimulus onset (the C1 component) that varies in polarity according to stimulus position, followed by a positive wave at 100–130 ms (P1 component) and a negative wave complex at 140–190 ms (N1 component). Conflicting proposals have been advanced regarding the neural generators of these components in striate and extrastriate cortical areas (reviewed in Clark et al., 1995). Recent studies that combined dipole modeling with functional magnetic resonance imaging (fMRI) of activated cortical areas have provided evidence that the C1 component to pattern onset is generated in the primary visual cortex (area V1) while the early part of the P1 wave arises from generators in the dorsal extrastriate occipital cortex and the later part of the P1 from sources in ventral extrastriate cortex (Di Russo et al., 2001, Di Russo et al., 2003, Martinez et al., 1999, Martinez et al., 2001). These studies localized the generators of the early, anterior part of the N1 complex to posterior parietal cortex and the later posteriorly distributed N1 to the same occipital generators as the P1.

The VEP elicited by a reversing pattern (e.g., checkerboard) is characterized by an initial negative component peaking at around 75 ms after the reversal (N75) followed by a positive component at around 100 ms (P100) and a second negativity at around 145 ms (N145) (Halliday, 1993). As with the pattern-onset VEP, the neural generators of the pattern-reversal VEP components have been much debated. While there is widespread agreement (based on studies in both animals and humans) that the N75 of the pattern-reversal VEP originates from the primary visual cortex (see Table 1), the origin of the P100 component is controversial. Some investigators have suggested that the P100 (like the pattern-onset P1) is mainly generated in extrastriate visual areas, while the majority has proposed that the P100 is generated (like the N75) in the striate cortex. N145 has been studied less extensively than the two earlier components. Some studies have identified a source for N145 in the extrastriate visual cortex, while others concluded that the N145 arises from the calcarine cortex or from both striate and extrastriate areas. Despite uncertainties as to the neural origins of its components, the pattern-shift VEP has had wide applications in clinical studies of visual system dysfunction (Halliday, 1993).

This lack of agreement among previous studies may be due to methodological differences such as the number of recording sites and the type of stimuli used. In particular, a sparse electrode array may not be able to differentiate concurrent activity arising from neighboring visual areas nor to obtain an accurate picture of the voltage topography produced by a given source. Furthermore, the use of stimuli extending over wide visual angles in some studies might have activated widespread regions of retinotopic cortical areas, thereby reducing the possibility of identifying the exact generator locations. In particular, stimuli that span more than one visual quadrant (crossing the horizontal and vertical meridians) may lead to activation of neuronal populations with opposing geometry (as in the primary visual area) resulting in a cancellation of their electric activity and in misinterpretation of the underlying source.

The purpose of the present study was to determine the detailed component structure of the pattern-reversal VEP and to localize its neural generators by using focal stimulation of each of the visual quadrants and a dense recording array of 64 electrodes. Sources were identified using dipole modeling based on a realistic head model and were compared with loci of cortical activations revealed by fMRI in response to the same stimuli. These sources were also localized on flat map with respect to visual cortical areas identified in individual subjects by retinotopic mapping and motion stimulation.