Elsevier

Clinical Neurophysiology

Volume 110, Issue 4, 1 April 1999, Pages 636-642
Clinical Neurophysiology

Gamma responses and ERPs in a visual classification task

https://doi.org/10.1016/S1388-2457(99)00002-4Get rights and content

Abstract

Objective: We examined event-related potentials (ERPs) and gamma range EEG activity in a visual classification task to assess which variables affect these responces.

Methods: Ten subjects silently counted the occurrence of rare Kanizsa squares (targets) among Kanizsa triangles and non-Kanizsa figures (standards). By applying a time-frequency analysis to the data and selectively calculating topographical maps of certain frequencies.

Results: We were able to find 3 different types of gamma responses to Kanizsa figures: an early phase-locked gamma response at 40 Hz in the N100 time range, late phase-locked gamma activity (200–300 ms) at 40 Hz and a continuous phase-locked gamma response at 80 Hz due to the monitor refresh frequency. The two 40 Hz responses were significantly higher for Kanizsa figures than for non-Kanizsa figures and within the Kanizsa figures were higher for the target figure than for the non-target.

Conclusion: The phase-locking of these two responses, previously found also as non-phase-locked activity, could be synchronized due to the monitor flicker frequency. Also, our findings suggest that the gamma responses are not solely associated with the binding of stimulus features, but reflect some processes related to target processing.

Introduction

Currently, many investigations in cognitive neuroscience tackle the aspects of brain activity associated with binding multiple features together to form one coherent object. Synchronization of neuronal responses could in principle achieve this task (von der Malsburg and Schneider, 1986, Singer, 1993). Frequencies in the gamma range (γ) are believed to give a clue to which cortical areas are involved in the representation of the binding processes, since they have been found in single-cell studies with animals (Gray and Singer, 1989, Engel et al., 1992), as well as in scalp-recorded EEG of human subjects (Tallon-Baudry et al., 1997). While Jürgens et al. (1995) argued that γ responses are sometimes merely harmonics of changes in the α frequency band (8–12 Hz), Pulvermüller et al. (1995) found γ responses that discriminated between words and psuedo-words and found no evidence of correlation to power changes outside the γ range. A review and classification of γ band activity demonstrates the psychophysiological correlates and functional significance of these responses (Başar-Eroglu et al., 1996). Such γ frequencies can be found phase-locked (Tallon et al., 1995) and non-phase-locked to the onset of experimental stimuli (Eckhorn et al., 1990; Tallon-Baudry et al., 1996). Psychophysical experiments have recently shown that flickering stimuli in the gamma range can enhance visual binding (Elliott and Müller, 1998), providing further evidence for the relevance of γ activity for binding processes. In that study, Kanizsa-like figures had to be detected and reaction times significantly decreased when the targets were pre-attentively flickering at a frequency of 40 Hz prior to the detection period, as compared to other flickering frequencies.

In our experiment, 10 subjects silently counted the occurrence of rare Kanizsa squares (targets) among 3 types of standards – Kanizsa triangles, triangular non-Kanizsa figures and rectangular non-Kanizsa figures. In order to detect γ activity, a time-frequency analysis based on a wavelet transform was employed. The wavelet approach for the time-frequency analysis of electrophysiological signals has proven very efficient (Schiff et al., 1994, Halgamuge et al., 1996, Herrmann and Reine, 1996, Herrmann, 1997). Recently, some researchers have applied the technique of wavelet transforms for viewing time-frequency details of ERPs (Tallon-Baudry et al., 1997). In addition to computing the time-frequency planes, we extracted frequency bands from it for subsequent generation of topographical maps for certain frequencies.

Section snippets

Subjects

Ten subjects with a mean age of 24.3 years (ranging from 20 to 35 years, 6 female) participated in the study. All subjects were right-handed and had normal or corrected-to-normal vision. They showed no signs of neurological or psychiatric disorders and all gave written informed consent.

Stimuli

As in the experiments of Tallon-Baudry et al. (1996), we used Kanizsa figures as stimulus material. According to Tallon-Baudry et al. (1996), it is advantageous to use a Kanizsa target which is different from the

Results

All of the subjects performed well on the detection of the Kanizsa squares. The mean error rate was 2.3% (SD 4.8%). One of the subjects showed strong technical artifacts in several electrodes and had to be excluded from further analysis.

Discussion

Our ERP results are consistent with the view that visual stimulus properties are being processed in two stages. In a first stage, purely physical stimulus properties like changes in illumination play the dominant role. This is reflected in the amplitude of the N20 (30–60 ms) which is higher for the squares as compared to the triangles. We argue that this is due to the fact that four pac-men, when flashed onto the screen, lead to a greater change in overall brightness than only three of them. In

References (29)

  • K. Chiappa

    Evoked potentials in clinical medicine

    (1997)
  • R. Eckhorn et al.

    Stimulus-related facilitation and synchronization among visual cortical areas: experiments and models

  • M. Elliott et al.

    Synchronous information presented in 40 Hz flicker enhances visual feature binding

    Psycholog Sci

    (1998)
  • C. Gray et al.

    Stimulus-specific neuronal oscillations in orientation columns of cat visual cortex

    Proc Natl Acad Sci USA

    (1989)
  • Cited by (215)

    • Haptic recognition memory following short-term visual deprivation: Behavioral and neural correlates from ERPs and alpha band oscillations

      2018, Biological Psychology
      Citation Excerpt :

      The function was applied for each electrode over an epoch of −1000 to 2000 ms with a resolution of 2 ms for each time point (500 Hz sampling rate). The wavelet had a central frequency = 1 and a bandwidth = 1 (Herrmann, Mecklinger, & Pfeifer, 1999; see Sebastian & Ballesteros, 2012, for a description of the method). The width was of 6.28 oscillations, over a total of 4560 epochs (2349 for the control group and 2211 for the experimental group).

    • Electrophysiological responses to symmetry presented in the left or in the right visual hemifield

      2017, Cortex
      Citation Excerpt :

      Interestingly, regularity did not consistently influence N1 amplitude in our study. It has previously been reported that N1 amplitude increases with gestalt like images (Brodeur et al., 2008; Herrmann & Bosch, 2001; Herrmann, Mecklinger, & Pfeifer, 1999). The N1 component is sometimes sensitive to symmetry (Makin et al., 2012; Makin et al., 2013) however this has not been found consistently (Höfel & Jacobsen, 2007a, 2007b; Jacobsen & Höfel, 2003; Norcia et al., 2002).

    View all citing articles on Scopus
    View full text