Elsevier

Cognitive Brain Research

Volume 20, Issue 3, August 2004, Pages 376-383
Cognitive Brain Research

Research report
Alpha rhythm of the EEG modulates visual detection performance in humans

https://doi.org/10.1016/j.cogbrainres.2004.03.009Get rights and content

Abstract

The effects of the changes in the frequency spectrum of the electroencephalogram (EEG) on the perception of near-threshold visual stimuli and on the event-related potentials (ERPs) produced by these stimuli were investigated on 12 healthy volunteers. The stimulus intensity, at which each subject could detect 50% of the presented stimuli, was defined as the sensory threshold for that subject. Single ERP trials were separated into two groups: trials with detected and undetected stimuli. The ERPs and the average power spectra of the 1 s prestimulus periods were computed for both conditions. P300 amplitudes of the ERPs, and total power and relative band powers of the delta (0.5–4 Hz), theta (4–7.5 Hz), alpha (7.5–13 Hz), beta (13–30 Hz), and gamma (30–70 Hz) frequency bands of the prestimulus power spectra were measured. Between the two conditions, a specific difference was observed in the relative power of the alpha band, which was significantly lower before detected stimuli (p<0.01) in line with significantly higher amplitudes of the ERPs (p<0.001). These results show that short-lasting changes in brain's excitability state are reflected the relative alpha power of the EEG, which may explain significant variability in perceptual processes and ERP generation especially at boundary conditions such as sensory threshold.

Introduction

Analyses of the electroencephalogram (EEG) and event-related potentials (ERPs) are among the fundamental research tools for understanding the sensory and cognitive functions of the brain. Many studies have shown the relationship between the electrical activity of the brain and complex psychophysiological processes such as alertness, arousal, attention, memory and executive functions [12], [28], [29], [33]. However, subject's ability to sustain the initial level of performance to tasks during electrophysiological recordings is limited, whereas, due to the low ERP/spontaneous EEG signal ratio, the measurement of the ERPs requires the repetition of the same event a number of times during each experimental condition and averaging of the time-locked EEG segments.

Makeig and Jung [24] reported that the performance fluctuations are accompanied by changes in the frequency spectrum of the EEG. According to Steriade et al. [38], such changes appear to index the action of brain mechanism involved in modulating arousal and/or sensory processing in and near to sleep. Although the effects of such large-scale EEG changes on the sensory processing are clearly demonstrated, the relationship between the short-lasting microchanges of the ongoing EEG and perceptual processes are less studied [2]. While some studies have demonstrated the relations between the prestimulus EEG and the event-related potential amplitude and latencies [2], [31], others have focused on the relationship between the spontaneous EEG and the cognitive performance of the subject [20], [22]. Experiments at or near sensory threshold may help to integrate both approaches by providing a task, where subject's performance is strongly dependent on microlevel arousal changes in contrast to large-scale differences observed near to sleep. Hence, perception of the low-intensity stimuli may build a good model to understand the relationship between the microlevel changes in the ongoing EEG and both the following ERP and the perceptual performance.

At low stimulus intensities, some of the presented stimuli are detected in contrast to others, which are missed. Therefore, the sensory threshold can only be defined in a stochastic manner as the stimulus intensity, at which 50% of the presented stimuli can be detected [13]. This stochastic property complicates the understanding of the neural mechanisms of sensory processing in general. Any EEG measure that determines whether the following near-threshold stimulus will be detected might be very important in reducing our stochastic view of the near-threshold perception and sensory processing mechanisms in general.

There are some studies showing ERP changes obtained either indirectly or directly in response to subliminal stimuli [6], [7], [9]. Hence, quantification of the electrophysiological responses to threshold-level sensory stimuli is possible.

Based on these facts, we previously studied the effects of spontaneous slow cortical potential shifts (SCP) in the occipital area on signal detection performance and showed that during slow negative shifts of the EEG a significantly higher number of presented low-intensity stimuli could be detected in comparison to the periods with slow positive EEG shifts [10].

The aim of the present study is to analyze in a similar manner the relations between the rhythmic EEG components and both the detection of the threshold-level visual stimuli and ERPs obtained with these stimuli.

Section snippets

Subjects

Twelve healthy volunteers (between 26 and 44 years, four males, eight females) participated in the study. Seven subjects had normal, and five had corrected to normal visual functions.

Recording procedures

Subjects were seated in an armchair in an electrically shielded, sound-attenuated room. EEG data were recorded from Ag–AgCl disc electrodes placed on Oz, O1, O2, Pz, P3, P4, Cz, C3, C4, Fz, F3 and F4 locations (10/20 system) against the reference of linked earlobes. All electrode impedances were less than 10 kΩ.

Results

The subjects could detect 121±25 out of 300 stimuli (40±8.3%). After artifact elimination, the mean number of trials with a detected stimulus across the subjects was 74±22. The number of trials, at which the subjects could not detect the stimulus, was equalized to the number of detected stimuli by randomly excluding some sweeps to avoid the effects of the different numbers of sweeps on the averaged ERP amplitude.

The grand averages of the ERPs (N=12) under both conditions (detected vs.

Alpha activity, visual detection performance around the sensory threshold, and ERPs

The results of the present study show that the detection threshold for visual stimuli depends significantly on the alpha power in the prestimulus EEG. With decreasing stimulus intensities, the subjective confidence on the detection of a stimulus decreases continually. Therefore, the sensory threshold is defined as the intensity, where the detection of the stimuli, hence confidence of the subject on the conscious perception of the stimuli is at chance level; that is, half of the trials are

Acknowledgements

This work is supported by The Research Fund of the Istanbul University project numbers: BYP-89/12112002, BYP-90/12112002, and 2002 TUBITAK-BAD Lundbeck grant.

References (38)

Cited by (0)

View full text