Abstract

Do ongoing brain states determine conscious perception of an upcoming stimulus? Using the high temporal resolution of EEG, we investigated the relationship between prestimulus neuronal oscillations and the perceptibility of two competing somatosensory stimuli embedded in a backward masking paradigm. We identified two prestimulus EEG signatures predictive for a suprathreshold yet weak target stimulus to become perceptually resistant against masking by a stronger distractor stimulus: (i) over left frontal cortex a desynchronization of the regional beta rhythm (∼20 Hz) 500 msec prior to a perceived target, and (ii) a subsequent additional attenuation of both mu (∼10 Hz) and beta “idling” rhythms over those pericentral sensorimotor cortices which are going to process the upcoming target stimulus. Furthermore, across subjects the probability for target perception strongly correlates with the individual absolute level of pre-target amplitudes in these frequency bands and locations. These signatures significantly differed from the EEG characteristics preceding detected and undetected single stimuli. We suggest that the early activation of left frontal areas involved in top–down attentional control is critical for preventing backward masking and leads the preparation of primary sensory cortices: The ensuing prestimulus suppression of sensory idling rhythms warrants an intensified poststimulus processing, and thus, effectively promotes conscious perception of suprathreshold target stimuli embedded into an ecologically relevant condition featuring competing environmental stimuli.

INTRODUCTION

A multitude of environmental events impinges on the receptor surfaces of our senses, yet only a minor fraction of them is perceived consciously. This raises intriguing questions about where and how in our brains the perceptual fate of a particular stimulus is determined. In the somatosensory domain, external factors such as stimulus intensity and stimulus competition are known to interact with, but not to fully determine internal selection processes: Lowering the stimulus intensity can decrease the probability of stimulus perception from unity to zero—yet, interestingly, single near-threshold stimuli of identical low intensity may either be perceived or go unnoticed. The source for this behavioral variability has been looked for at the cortical level but sensory-specific cortices have been found activated following both outcomes: Intracranial recordings in monkeys (de Lafuente & Romo, 2005) and humans (Ray et al., 1999; Libet, Alberts, Wright, & Feinstein, 1967) as well as a recent combined MEG/computational modeling study (Jones, Pritchett, Stufflebeam, Hämäläinen, & Moore, 2007) revealed that early neuronal activity in primary somatosensory cortex (S1), which reflects thalamocortical input and the ensuing primary cortical response, is evoked independently of whether or not such threshold stimuli were detected. Remarkably, even if stimuli are delivered at clearly suprathreshold intensities, yet followed by a competing stimulus, poststimulus S1 activation does not warrant conscious perception: Studies with patients who suffer from unilateral somatosensory extinction during bilateral stimulation revealed activation of S1 in fMRI as well as in early somatosensory evoked potentials (SEPs) for both perceived and unperceived suprathreshold target stimuli (Sarri, Blankenburg, & Driver, 2006; Eimer, Maravita, Van Velzen, Husain, & Driver, 2002). This finding was recently generalized beyond brain lesion studies in an SEP study mimicking extinction in healthy subjects (Schubert, Blankenburg, Lemm, Villringer, & Curio, 2006).

Apart from these reports on poststimulus processing, only a few studies have dealt with the complementary quest for brain states preceding the arrival of a stimulus which are predictive of its further perceptual destiny. Convergent studies in both the visual and somatosensory system showed that the detectability of near-threshold stimuli is modulated by EEG and MEG rhythms at about 10 Hz (alpha) prevailing in the respective modality-specific cortices immediately prior to stimulus delivery (van Dijk, Schoffelen, Oostenveld, & Jensen, 2008; Yamagishi, Callan, Anderson, & Kawato, 2008; Hanslmayr et al., 2007; Thut, Nietzel, Brandt, & Pascual-Leone, 2006; Ergenoglu et al., 2004). Hanslmayr et al. (2007) divided their subjects in perceivers and nonperceivers (in relation to the individual perception rate of visual stimuli) and found that perceivers showed a significantly lower occipital alpha power than nonperceivers. Furthermore, they found a negative correlation between perception rate and alpha power for the group of perceivers. Applying a task of visuospatial attention, Yamagishi et al. (2008) and Thut et al. (2006) demonstrated a decrease in alpha power in the hemisphere contralateral to the focus of attention preceding a performance increase for subsequent stimuli in the attended hemifield. The authors concluded that the covert shift of attention changes the sensitivity of early visual areas expressed in a decrease of alpha power. This idea was recently corroborated by a study applying transcranial magnetic stimulation (TMS) (Romei, Rihs, Brodbeck, & Thut, 2008; Romei et al., 2007). Romei and colleagues used TMS to induce illusory visual percepts (phosphenes) in blindfolded participants while recording EEG. They showed that during periods of lower alpha amplitudes, phosphenes were more easily evoked than during higher alpha. This finding clearly argues for the idea that different levels of occipital alpha power represent different levels of excitability of this region and low alpha amplitudes signify high excitability. A recent MEG study (van Dijk et al., 2008) investigated the relation between the power of prestimulus occipital rhythmic alpha activity and visual discrimination ability by presenting visual stimuli with subtle differences in gray levels which subjects were asked to respond to when detected. When the color differences were individually adjusted at a 50% detection threshold, prestimulus occipital alpha band power was found to correlate with the detection of this just noticeable difference in visual contrast: The subject's behavioral discrimination capability decreased with increasing prestimulus alpha power, possibly related to the concept that an increase in occipital alpha power reflects inhibition of posterior cortices (for a review, see Klimesch, Sauseng, & Hanslmayr, 2007), hindering the transmission of information to higher cortical areas and thereby impairing conscious perception. In the somatosensory system, perception and timing of the corresponding behavioral response to near-threshold somatosensory stimuli have been found optimal if preceded by a medium-sized amplitude of contralateral sensorimotor rhythms (Linkenkaer-Hansen, Nikulin, Palva, Ilmoniemi, & Palva, 2004) or by an increase in their phase-locking (Palva, Linkenkaer-Hansen, Näätänen, & Palva, 2005).

Building on these findings, we aimed here to broaden the notion of a possibly causal EEG-to-perception link: To this end, we investigated those cortical processes that enable a weak, yet clearly suprathreshold, stimulus to withstand masking by a second stronger stimulus. Thereby, we modeled the more general and ecologically valid case of multiple and suprathreshold stimuli which compete for perception, representing a common environmental situation, and compared the results to perceptual effects preceding single stimuli. We used data from a paradigm mimicking extinction following electrical stimulation of the left and right index finger in healthy subjects. These data split into masked versus perceived target trials have allowed us to identify components of averaged poststimulus SEP specifically related to conscious perception (Schubert et al., 2006). In the present study, we analyzed EEG raw data in the time–frequency domain to investigate the possible influence of prestimulus sensorimotor rhythms on masking efficiency.

METHODS

Subjects and Recordings

EEG data have been recorded from 12 right-handed subjects (24–33 years, 2 women) with 62 Ag/AgCl electrodes (58 EEG electrodes in the 10–20 system and 4 EOG electrodes), referenced against the tip of the nose. Horizontal EOG (hEOG) was recorded bipolarly from the outer canthi of both eyes and vertical EOG (vEOG) was recorded above and below the left eye. The data were digitized at 5 kHz. The study was approved by the local Ethics Committee of the Charité, University Medicine Berlin.

Procedure and Experimental Paradigm

Right and left index fingers were stimulated each with a constant current pulse (0.5 msec duration, DS7A, Digitimer, UK) applied via adhesive electrodes to the finger tip. At the beginning of the experiment, sensory thresholds were determined separately for each index finger by the method of limits, applying five repetitions of increasing and decreasing stimulus intensities. The stimulus intensity delivered to the right index finger (mask stimulus) was adjusted just below the subject's individual pain threshold. The stimulus intensity applied to the left index finger (target stimulus) was set to 30–40% above the individual's sensory threshold. To compensate for potential threshold drifts, this procedure was repeated after every fourth block during the experiment.

During experimental blocks, the subjects fixated a cross displayed at the center of a 15-in. computer screen placed in front of them. The experiment consisted of 12 blocks, each containing 128 stimulation trials. In 50% of trials, left targets were followed by right mask stimuli with an interstimulus interval of 70 msec found previously to ensure a most effective backward masking of the target stimuli (Meador, Ray, Day, Ghelani, & Loring, 1998). Twenty-five percent of trials contained a single-left stimulus, and the remaining 25% were single-right trials. The intertrial interval (ITI) varied randomly in steps of 500 msec between 2000 and 3500 msec. Additionally, stimulus onsets were jittered between 0 and 100 msec in steps of 1 msec to avoid locking on power-line interferences. Subjects were instructed to focus their attention on the left index finger and to press a left foot pedal with their right foot whenever they perceived the weak left target, irrespective of whether it was presented alone or followed by the strong right mask, and to press a right foot pedal with their right foot whenever they perceived the mask only.

To minimize any potentially interfering effects which could be held responsible instead of masking when subjects did not perceive the left target stimulus in the double-stimulation condition, an a priori exclusion criterion was defined: An experimental block was included for further data analysis only if the target detection rate in single-left trials was significantly higher than in double-stimulation trials (χ2-frequency distribution). Due to this criterion, 75 of the 144 blocks in total had to be excluded and the resulting mean number of resulting trials was 116 for detected targets (SD = 46) and 250 trials for masked targets (SD = 116). Behaviorally, the target detection rate declined from 75% (single-left trials, i.e., targets presented without mask) to 32% (double-stimulation; targets followed by a mask). Mean reaction time was 797 msec (SD = 96 msec) after stimulus presentation (target = 794 msec; mask = 806 msec; difference ns).

Data Analysis

In the present study, we analyzed EEG data from both single- and double-stimulation trials. Off-line, the data were low-pass filtered to 100 Hz by applying a fifth-order Butterworth filter, downsampled to 500 Hz, and re-referenced against a common average reference. We epoched the data relative to the onset of the target stimuli and conducted a time–frequency analysis by convolving each epoch with complex Morlet wavelets wf (t) = exp(2iπft)((2π)0.5fω0−1 exp(−2(πft)2ω0−2)), that is, sinusoidals being localized in time by a Gaussian envelope. Standard deviations (widths) of the wavelet are given by σ(f) = ω0/2πf in time domain and η(f) = f0 in frequency domain, where the parameter ω0 was set to 5.5 for all wavelets. Wavelet analysis was carried out both for the pre- and the poststimulus intervals. Importantly, we ensured that no information from the “wrong” side of the stimulus could affect the result, as for each analysis we mirrored the respective part of the EEG at the stimulus onset.

To characterize the topographic distribution of potential EEG differences between trials with perceived and unperceived targets, we performed a wavelet analysis in the mu and beta band for 58 scalp electrodes (except hEOG and vEOG). The data analysis was restricted to these bands because their signal-to-noise ratio permitted to reliably extract the signals given the available recording settings; exploratory analyses had shown that the gamma band was compromised by high-frequency head muscle artifacts, and the theta–delta band by movement artifacts. Importantly, absolute amplitude values of the mu and beta band were taken for the analysis of the data. For the mu band, three wavelets with center frequencies between 10.4 and 14.9 Hz were used. Another four wavelets were selected for the beta band (center frequencies between 17.9 and 25.8 Hz). The frequencies were chosen to establish a gap of approximately one frequency-domain standard deviation between both bands. The EEG data were convolved using Zfj(t) = X(t) × Wfj(t), where the filters Wfj(t) were constructed by discretizing the wavelets wfj(t) between −2σ(fj) and 2σ(fj). We computed (logarithmized) amplitudes in each of the wavelet-filtered sequences and calculated the average in an interval immediately before (−400 msec to 0 msec) or after (0 msec to 500 msec) the stimulus and across the frequencies of each band; thereby, we obtained one value for each electrode and band.

Because the main purpose of this study was to investigate prestimulus brain states conducive for masking, we further related the topography of the obtained prestimulus differences to the loci of poststimulus event-related desynchronization (ERD) (for a review, see Pfurtscheller & Lopes da Silva, 1999), which can be utilized as a functional landmark indicating the hand region of primary sensorimotor cortices. Significant ERDs in the beta band [p < .05; Bonferroni corrected (corr.) for the number of electrodes (58)] were mapped for trials with detected single left-hand or, respectively, single right-hand stimuli.

In addition to the high-resolution topographical analysis, we investigated the time course of spectral changes at the electrode carrying the maximal effect in the spatial analysis. To this end, we kept 25 samples of each of the wavelet-filtered sequences between −1000 msec and 0 msec and 37 samples between 0 msec and 1500 msec for pre- and poststimulus analyses, respectively. From these data, grand-average ERD curves were calculated.

To complement the analysis of prestimulus induced masking, we investigated the dependence of both topographical and temporal predictors on the ITI. Specifically, we repeated the abovementioned analyses for subgroups of trials preceded by either short (2000 msec) or long (3500 msec) ITIs.

To test the statistical significance of the differences between the amplitudes of perceived and masked stimuli, we performed subject-wise two-sample t tests for the topographical and time course analyses. The obtained t scores were transformed into z scores of a standard normal distribution and a one-sample z test was performed across the scores of all subjects. We used Bonferroni-corrected p values for the analysis of the topographic distribution based on the number of electrodes (58), and for the temporal analysis based on the number of time bins (25 and 37, respectively). Note that the statistical approach taken here accounts for the individual amplitude of each subject. For this reason, slight discrepancies between the test statistics and the grand-average difference might be observed in Figures 1 and 2.

Figure 1. 

(A) Topographies: Significant EEG amplitude differences for perceived compared to masked target stimuli shown here for the grand-average scalp topographies determined prior to stimulus delivery (averaged from −400 msec to target onset) in both the mu band (10.4–14.9 Hz) and the beta band (17.9–25.8 Hz). Red color denotes areas with amplitudes significantly larger for masked target stimuli [corrected p values for the number of electrodes (58)]. Blue lines in the lower graph delineate the distribution of poststimulus beta desynchronizations for left-hand (right hemisphere) and right-hand (left hemisphere) single stimuli (p < .05), serving as functional landmark for the respective primary sensorimotor cortices. Time courses: Grand-average time course of mu and beta amplitude changes at electrode C4 and FC3. For each band, the mean amplitudes from −1000 msec to target onset for perceived and unperceived targets are shown (mean ± 1 SEM). Colors along the x axis show statistical significance of the differences between the amplitudes from a subject-wise two-sample t test. The obtained t scores were transformed into z scores of a standard normal distribution and a one-sample z test was performed across the scores of all subjects. (B) Grand-average time–frequency plots for the prestimulus time interval from −1000 msec to target onset showing the difference between to-be-perceived and to-be-masked target stimuli. Red color denotes areas with amplitudes significantly [corrected for the number of time bins (25)] larger for masked target stimuli and blue larger for perceived targets. Plots were calculated for those electrodes showing the strongest effects in the topographic analysis (FC3, C4).

Figure 1. 

(A) Topographies: Significant EEG amplitude differences for perceived compared to masked target stimuli shown here for the grand-average scalp topographies determined prior to stimulus delivery (averaged from −400 msec to target onset) in both the mu band (10.4–14.9 Hz) and the beta band (17.9–25.8 Hz). Red color denotes areas with amplitudes significantly larger for masked target stimuli [corrected p values for the number of electrodes (58)]. Blue lines in the lower graph delineate the distribution of poststimulus beta desynchronizations for left-hand (right hemisphere) and right-hand (left hemisphere) single stimuli (p < .05), serving as functional landmark for the respective primary sensorimotor cortices. Time courses: Grand-average time course of mu and beta amplitude changes at electrode C4 and FC3. For each band, the mean amplitudes from −1000 msec to target onset for perceived and unperceived targets are shown (mean ± 1 SEM). Colors along the x axis show statistical significance of the differences between the amplitudes from a subject-wise two-sample t test. The obtained t scores were transformed into z scores of a standard normal distribution and a one-sample z test was performed across the scores of all subjects. (B) Grand-average time–frequency plots for the prestimulus time interval from −1000 msec to target onset showing the difference between to-be-perceived and to-be-masked target stimuli. Red color denotes areas with amplitudes significantly [corrected for the number of time bins (25)] larger for masked target stimuli and blue larger for perceived targets. Plots were calculated for those electrodes showing the strongest effects in the topographic analysis (FC3, C4).

Figure 2. 

Topographies: Significant EEG amplitude differences prior to (averaged from −400 msec to target onset) perceived compared to masked/missed target stimuli (A and C) or following double stimulation (averaged from target onset to 500 msec, B) in both the mu band (10.4–14.9 Hz) and the beta band (17.9–25.8 Hz). Red color denotes areas with rhythm amplitudes significantly larger for unperceived target stimuli and vice versa for blue color [corrected p values for the number of electrodes (58)]. Time courses: Grand-average time course of mu and beta amplitude changes at the electrode showing the maximum difference in topographies. For each band, the mean amplitudes from −1000 msec to target onset (A and C) and from target onset to 1500 msec for perceived and unperceived targets (B) are shown (mean ± 1 SEM; corrected p values for the number of time bins; prestimulus: 25; poststimulus: 37). Colors along the x-axes show statistical significance of the differences between the amplitudes from a subject-wise two-sample t test. The obtained t scores were transformed into z scores of a standard normal distribution and a one-sample z test was performed across the scores of all subjects. (A) Data were analyzed separately for trials following a short intertrial interval (ITI, 2000 msec) or long ITI (3500 msec). (B) Poststimulus difference for perceived versus masked double-stimulation trials. (C) Prestimulus difference for perceived versus missed single target stimuli.

Figure 2. 

Topographies: Significant EEG amplitude differences prior to (averaged from −400 msec to target onset) perceived compared to masked/missed target stimuli (A and C) or following double stimulation (averaged from target onset to 500 msec, B) in both the mu band (10.4–14.9 Hz) and the beta band (17.9–25.8 Hz). Red color denotes areas with rhythm amplitudes significantly larger for unperceived target stimuli and vice versa for blue color [corrected p values for the number of electrodes (58)]. Time courses: Grand-average time course of mu and beta amplitude changes at the electrode showing the maximum difference in topographies. For each band, the mean amplitudes from −1000 msec to target onset (A and C) and from target onset to 1500 msec for perceived and unperceived targets (B) are shown (mean ± 1 SEM; corrected p values for the number of time bins; prestimulus: 25; poststimulus: 37). Colors along the x-axes show statistical significance of the differences between the amplitudes from a subject-wise two-sample t test. The obtained t scores were transformed into z scores of a standard normal distribution and a one-sample z test was performed across the scores of all subjects. (A) Data were analyzed separately for trials following a short intertrial interval (ITI, 2000 msec) or long ITI (3500 msec). (B) Poststimulus difference for perceived versus masked double-stimulation trials. (C) Prestimulus difference for perceived versus missed single target stimuli.

Correlation Analysis

A correlation analysis was run between the absolute individual averaged mu and beta amplitude values preceding to-be-detected target stimuli and the perception rates for left target stimuli when followed by a mask. To this end, we calculated the Spearman correlation coefficient across subjects between amplitudes in these bands (from −400 msec to stimulus onset from the electrodes showing the maximum effects) and the perception rate (number of perceived divided by the number of all double stimulation trials).

RESULTS

Prestimulus Effects of Perceptual Masking

Topographical Analysis

Prior to masked left-hand target stimuli, the oscillation amplitudes in mu and beta bands were significantly larger over sensorimotor cortices of the left hand (Figure 1A; central electrodes of right hemisphere: mu band, p < .01, corr.; beta band, p < .01, corr.). Furthermore, preceding perceived target stimuli, significantly lower amplitudes were detected in the beta band over left-frontal cortex (p < .05, corr.). Over the right hemisphere, the location of the prestimulus beta band amplitude difference was fully overlapping with the location of the poststimulus ERD, an attenuation of the resting rhythm at sensorimotor hand cortex induced by single left-hand stimuli. Over the left hemisphere, the perception-related prestimulus beta band differences were located distinctly more frontal than the sensorimotor hand area as delineated by the poststimulus beta ERD induced by single right-hand stimuli.

For the subaverages of long ITI trials, the results from the overall average were confirmed, that is, perceived targets were preceded by a desynchronization of the mu and beta rhythmic activity in contralateral sensorimotor cortex (mu: C4, p < .01, corr. and a trend for the effect in the beta band p < .10, corr.; Figure 2A). Interestingly, for short ITIs, no significant modulation of the mu band preceding perceived compared to masked targets was revealed. Here, a significant effect was found only for the beta band, showing a significant amplitude enhancement for perceived compared to masked stimuli at centro-parietal positions (maximal at CPz; p < .01, corr.).

Time Courses

To functionally interpret the topographies of prestimulus EEG rhythm differences, we analyzed the prestimulus time course of mu and beta amplitude changes for perceived versus masked target stimuli. For double stimulation, we found significant amplitude differences starting already about 500 msec before target onset (Figure 1A): Over primary sensorimotor cortices of the left hand (representative electrode: C4), mu band amplitudes were significantly larger prior to masked targets starting at −340 msec (p < .05, corr.), and beta band amplitudes were larger starting at −420 msec (p < .05, corr.). For the left frontal prestimulus difference (electrode FC3), significantly smaller beta band amplitudes preceding to-be-perceived targets were detected as early as −500 msec (p < .05, corr.). This frontal beta rhythm difference stayed significant for 120 msec. These exemplary results are seen in full context in a time–frequency analysis (Figure 1B): For effective masking, at about 500 msec prior to target presentation, significant higher amplitudes emerge in the frequency range above 18 Hz over left frontal cortex and above 10 Hz over sensorimotor cortex.

For short ITIs, although there is a trend for a continuous enhancement for perceived targets in the centro-parietal beta band starting at −1000 msec (representative electrode: CPz), the effect becomes significant only at 380 msec before stimulation onset (p < .05, corr.; Figure 2A).

For long ITIs, a significant decrease of the mu band amplitudes over primary sensorimotor cortices of the left hand (representative electrode: C4), prior to perceived targets, also emerges at 380 msec before presentation of the target (p < .05, corr.). The time course of the beta band amplitudes at the same location also shows a clear decrease 400 msec prior to perceived targets, significant between −400 and −380 msec (p < .05), and a trend (p < .10) from −420 to 400 msec and from −220 to −180 msec. Although not significant in the topographic analysis, we analyzed the difference in the beta band at electrode FC3 to show the correspondence with the result for the averaged ITI. Here a very early significant difference expressed in lower amplitudes for subsequent perceived targets was found from −820 to −780 msec (p < 0.05, corr.) and a trend can be seen from −520 to −420 msec as well as from −320 to −180 msec before stimulus onset (Figure 2A).

Interestingly, when comparing the behavioral responses for long versus short ITIs preceding the double stimulation trials, we found a significantly higher target perception rate, that is, resistance to masking, for long (37.3%) compared to short (29.6%) ITIs (p < .05).

Poststimulus Effects of Conscious Perception

Topographical Analysis

Following double stimulation (Figure 2B), a widely distributed amplitude suppression covering whole right posterior cortex in the mu and beta bands (p < .01, corr.) characterizes the successful detection of the target stimulus. For the beta band, the difference between masked and perceived stimuli also expands to contralateral frontal areas and to a smaller extent to ipsilateral frontal cortex.

Time Courses

At 240 msec after application of the target stimulus, the amplitude in the mu band significantly decreases (representative electrode CP4, p < .01, corr.) for perceived compared to masked stimuli and remains significantly lowered up to 940 msec after target onset (Figure 2B). For the beta band, the contralateral effect emerges at 260 msec (representative electrode FC4, p < .05, corr. and p < .01, corr.) and remains significantly lowered up to 760 msec after target onset. For the ipsilateral hemisphere, the effects are not as prominent as in the contralateral hemisphere, but still significant (from 260 to 300 msec and from 500 to 740 msec at representative electrode FC3; p < .01, corr.).

Prestimulus Effects on Detection of Weak Suprathreshold Single Targets

The prestimulus effects for perceived single-left target stimuli strongly differ from the double-stimulation effects (Figure 2C). For the mu and beta bands, perceived targets are preceded by an increase of the amplitudes at centro-parietal electrode positions [p < .01 corr., for the mu band (at Pz) and p < .01 corr., for the beta band (at CPz)]. In the beta band, an additional fronto-central desynchronization, which is stronger over the right than the left hemisphere, predicts successful detection of a weak stimulus (p < .01 corr., at C4 and FC4).

Prior to single stimulation, the centro-parietal synchronization (at electrode Pz in the mu and at electrode CPz in the beta band) preceding perceived targets is significantly enhanced starting from −1000 msec before the onset (p < .05, corr.) continuing across the whole epoch (varying between ns and p < .01, corr.). In contrast, for the fronto-central beta desynchronization effect, amplitudes differ significantly starting at −240 msec (p < .05, corr.).

Correlation of Prestimulus EEG and Probability of Stimulus Perception

The correlation analysis between the absolute individual averaged mu and beta amplitude values preceding to-be-detected target stimuli and the perception rates for left target stimuli when followed by a mask revealed a significant negative correlation between the mu amplitude at C4 and perception rate (r = −.60, p < .05; Figure 3A). The negative correlation between the beta band amplitudes at C4 and the perception rate showed a trend (r = −.52, p = .084; Figure 3B) and was significant between beta band amplitudes at FC3 and perception rate (r = −.62, p < .05; Figure 3C).

Figure 3. 

Relation between the individual pre-target mu/beta rhythmic activity (averaged from −400 msec to target onset) and detection probability of the target stimulus. The three graphs show negative correlations between the amplitudes in these bands and the perception rates across the 12 subjects (dots) for (A) the mu band at electrode C4 and (B) for the beta band at electrode C4 and (C) FC3 (for details, see Methods). The graphs show that for both cortical locations and frequency bands, the target perception rate critically depends on the level of the prestimulus mu/beta amplitudes. r = correlation coefficient; p = p value.

Figure 3. 

Relation between the individual pre-target mu/beta rhythmic activity (averaged from −400 msec to target onset) and detection probability of the target stimulus. The three graphs show negative correlations between the amplitudes in these bands and the perception rates across the 12 subjects (dots) for (A) the mu band at electrode C4 and (B) for the beta band at electrode C4 and (C) FC3 (for details, see Methods). The graphs show that for both cortical locations and frequency bands, the target perception rate critically depends on the level of the prestimulus mu/beta amplitudes. r = correlation coefficient; p = p value.

DISCUSSION

Prestimulus Desynchronization of Rhythmic Activity of Primary Somatosensory Cortex

Are there characteristic brain states that can prevent a suprathreshold target stimulus from being masked, that is, favoring conscious target perception in an environment of competing stimuli? The present analysis of ongoing EEG oscillations revealed that a sensory-specific topographic pattern of lower mu and beta band amplitudes predicts a higher probability for an upcoming somatosensory target stimulus to be perceived in face of a competing stimulus: These effects related to prospective conscious perception were observed over the target-receiving primary sensorimotor hand area as indexed by a co-localization with the poststimulus beta ERD.

EEG studies employing near-threshold visual stimuli (van Dijk et al., 2008; Yamagishi et al., 2008; Hanslmayr et al., 2007; Thut et al., 2006; Ergenoglu et al., 2004) report a lower prestimulus occipital alpha power for perceived than for unperceived stimuli. The present result permits to generalize this notion as it establishes the prestimulus decrease of local mu/beta amplitudes over a primary sensory area as a common mechanism which significantly raises the probability for consciously perceiving upcoming stimuli not only in a near-threshold but also in a suprathreshold intensity regime. This is in line with the idea of desynchronization of low-frequency oscillations playing a constructive role for perception, for instance, in “priming” sensory areas to optimally process an expected stimulus (Engel, Fries, & Singer, 2001; Fries, Reynolds, Rorie, & Desimone, 2001). Thus, suprathreshold intensity, which could have prevented targets from being masked by a competing stimulus, cannot warrant target perception on its own; rather, an eventual resistance to masking requires a specific priming of receiving primary sensory cortex.

The present results may appear at variance with a recent EEG study applying a visual masking paradigm (Babiloni, Vecchio, Bultrini, Luca, & Rossini, 2006): For perceived stimuli, prestimulus alpha power was found increased at frontal and parieto-occipital sites. Only at the vertex electrode was the power reduced prior to perceived visual stimuli. This result was interpreted in relation to previous studies showing that good cognitive performance, for instance, in encoding and retrieval tasks, can be predicted by higher power or a tonic increase of prestimulus alpha power (Klimesch, Sauseng, & Gerloff, 2003; Klimesch, 1999). However, a direct comparison to the present study is not possible mainly because Babiloni et al. (2006) (i) calculate statistics of activity related to seen versus unseen stimuli with the outcome of a LORETA source modeling and not with the effects at electrodes; (ii) calculated prestimulus ERD relative to poststimulus ERD, each averaged over 1 sec; and (iii) combined their task of conscious perception, in which subjects had to respond orally with a task of subliminal priming requiring a manual keypressing, which favors interaction of multiple cognitive processes which had not been addressed in our study.

In a visual–spatial cueing paradigm applying target stimuli at a cued hemifield and distractor stimuli at an uncued location (Worden, Foxe, Wang, & Simpson, 2000), an increase of occipital alpha power was found in the hemisphere where processing of the distractor stimuli was anticipated, whereas alpha decreased in the hemisphere expected to process the target. Our results did not reveal an increase of the mu and beta rhythms over somatosensory cortex where the mask stimulus was anticipated. This discrepancy may be ascribed to differences in the applied paradigm as well as modality-specific stimulus processing: Although Worden et al. (2000) used trial-to-trial spatial cueing, implying transient attentional control, the subjects in our study were instructed to focus attention to the left index finger throughout the whole experiment, which represents a task of sustained attention. The cortical activation patterns related to transient and sustained top–down directed attention have been shown to differ markedly (Eimer & Forster, 2003). Furthermore, the lateralization of the somatosensory system may not be as distinct as for visual cortex. Only the first 50–80 msec poststimulus is assumed to constitute primarily contralateral somatosensory processing (e.g., Zhu, Disbrow, Zumer, McGonigle, & Nagarajan, 2007) Thus, weak ipsilateral somatosensory processing might diminish the effect of alpha power increase.

Hitherto, influences of prestimulus mu and beta rhythms on perception of somatosensory input have been addressed only for near-threshold single stimuli (Linkenkaer-Hansen et al., 2004) showing a nonlinear (inverted-U) relation between detection performance and rhythm amplitude over sensorimotor cortices. The present masking paradigm is different in four essential aspects. First, stimulation with single near-threshold stimuli may induce considerable changes in vigilance due to prolonged time periods during which the subjects have the impression of receiving no stimulation at all. In the present study, subjects received additional strong stimuli at the nontarget hand in 75% of all trials, thereby ensuring a continuous stream of stimuli with alerting intensities. Second, we included only trial blocks with a high detection rate of single weak (but still suprathreshold) stimuli in the analysis so that vigilance decrements cannot be held responsible for the high miss rate of weak stimuli in the double-stimulation trials in these blocks. Third, the present paradigm of bilateral masking involves additional cognitive processes such as spatial attention and inhibition of distractor interference. Thus, the differences between the results by Linkenkaer-Hansen et al. (2004) and the present study suggest different system states optimal for the detection of single near-threshold single stimuli and double suprathreshold target/mask stimuli. Fourth and perhaps the most important, these authors applied a different methodological approach to relate mu and beta amplitudes with performance: They divided their averaged MEG data into 10 percentiles dependent on the amplitude size of mu/beta activity. These percentiles were then correlated with the corresponding hit rates. This approach allows for a strong impact of the maximum and minimum values. As can be seen in their figure, the U-shape relation between mu amplitude and perception is driven by the 1st, 9th, and 10th percentile, meaning that at the extreme ends of the amplitudes, performance is very bad, whereas good performance is predicted by medium-sized amplitudes. The difference between this study and our results is that we split the EEG data relative to the behavioral outcome (perceived/masked) and then tested for statistical significance, which allows only for a smaller impact of the extreme values. Furthermore, for the beta band values in the study by Linkenkaer-Hansen et al., the U-shape is driven by the first percentile, otherwise, it would follow the negative correlation between beta amplitude size and perception rate as in our data.

Conscious Perception, Anticipatory Attention, and Enhanced Signal Processing

Conscious perception of a target stimulus, which could undergo masking by the co-occurrence of a second strong stimulus, crucially depends on the lowering of local “idling” rhythmic activity before the stimulus reaches primary sensory cortices. Smaller mu and beta amplitudes in the stimulus receiving somatosensory cortex may indicate an increase of its cellular excitability (Steriade & Llinas, 1988). A decrease of the 10-Hz rhythmic EEG/MEG activity over primary sensory cortices was also reported by Bastiaansen and Brunia (2001), investigating anticipatory attention to visual, somatosensory, and auditory stimuli. The authors suggest a neurophysiologic model of anticipatory attention where activation of sensory cortex in preparation to receive and optimally process an upcoming stimulus is indexed by a 10-Hz desynchronization.

In the current study employing bilateral masking, we suggest a specific relation between conscious stimulus processing and anticipatory attention to a spatial location: Preventing “extinction” may depend on the subjects' ability to selectively maintain their attentional focus on their left target finger while withstanding the potentially interruptive influence of repeated strong right-hand stimuli. Specifically, the instruction (“to focus attention on the left index finger in order to detect the weak target”) explicitly involved a task demand for spatial attention. Furthermore, increased poststimulus SEP components as observed for the parietal P100 and the frontal N140 for detected targets (Schubert et al., 2006) is frequently found in studies of somatosensory spatial selective attention (Forster & Eimer, 2004; Eimer, Forster, & van Velzen, 2003; Josiassen, Shagass, Roemer, Slepner, & Czartorysky, 1990; Michie, Bearpark, Crawford, & Glue, 1987) and can be integrated into the notion of attention as a necessary prerequisite for conscious perception (Dehaene & Naccache, 2001). We assume that prestimulus anticipatory attention to the left index finger favors ongoing processing in right somatosensory cortex by dampening its 10-Hz oscillatory activity. This results in a stronger response to the upcoming target stimulus at latencies more than 100 msec, preventing it from being masked by a competing stimulus, and thus, allowing for further processing leading eventually to conscious perception.

Prestimulus Synchronization and Desynchronization of Beta Band Oscillations

We observed considerable modulations of the beta frequency band prior to stimulation not only at the sensorimotor cortices but also at frontal and centro-parietal regions.

Sensorimotor Beta Desynchronization

In addition to studies investigating prestimulus effects of visual perception, Linkenkaer-Hansen et al. (2004) and the present study show a prestimulus modulation not only of the 10-Hz (alpha/mu) activity but also of rhythmic oscillations in the beta frequency band (∼15–30 Hz). Although the source of the mu rhythm has been ascribed to primary somatosensory cortex, the beta rhythm has been shown to be evoked more anteriorly in the primary motor area (Salmelin & Hari, 1994; Tiihonen, Hari, & Hämäläinen, 1989). This frequency band, together with the mu rhythm, also known as the Rolandic rhythms, both decrease in the contralateral hemisphere in response to sensorimotor activation, as in voluntary movement and somatosensory stimulation (Salenius, Schnitzler, Salmelin, Jousmäki, & Hari, 1997; Pfurtscheller, 1981). In the underlying study as well as in the study by Linkenkaer-Hansen et al., the somatosensory beta rhythmic activity shows a prestimulus modulation comparable to that of the mu rhythm, as expected from previous studies revealing similar modulation of both Rolandic rhythms.

Left Frontal Beta Desynchronization

We also identified a transient desynchronization of beta rhythms preceding perceived targets over left frontal cortex anterior to the sensorimotor hand area. The high temporal resolution of the EEG enables the detection of such short latency effects. As this effect was resistant also against the multiple comparison correction of p values (with respect to the number of time bins and electrodes), the possibility that it refers to an artifact is highly unlikely. Specifically, the frontal beta ERD started prior to the modulations at primary sensorimotor cortices, indicating a leading functional role for frontal vis-à-vis sensorimotor cortices. Although brain states preceding conscious perception have been subject to several recent studies (van Dijk et al., 2008; Yamagishi et al., 2008; Hanslmayr et al., 2007; Babiloni et al., 2006; Thut et al., 2006; Ergenoglu et al., 2004; Linkenkaer-Hansen et al., 2004), they did not report effects in the beta band in left frontal areas. This may have several reasons: Some authors restricted their analysis to sensors at primary sensory or centro-parietal areas (Romei et al., 2007, 2008; Yamagishi et al., 2008; Linkenkaer-Hansen et al., 2004). Others restricted their analysis to the alpha or lower frequency band (van Dijk et al., 2008; Babiloni et al., 2006; Thut et al., 2006). The fact that frontal beta desynchronization was not found in studies allowing for whole head and broad frequency band investigation of prestimulus effects on perception (Hanslmayr et al., 2007; Ergenoglu et al., 2004) may be attributed to the fact that these studies had used tasks involving different operant cognitive processes. We assume that the left frontal beta desynchronization may inherently be related to directing spatial attention to the locus of an anticipated stimulus. Furthermore, in case of near-threshold stimulation (50% detection rate) as used in the abovementioned studies, subjects may suffer from fluctuations of vigilance as they might not perceive stimulation for long periods, which interacts with perception. Indeed, in studies applying a warning cue or intermingled strong and even painful stimuli which both keep up a constant level of alertness, prestimulus frontal effects have been found in a series of recent findings: An fMRI activation of lateral frontal, and additionally, also parietal regions was found preceding the detection of low-intensity somatosensory stimuli while, during the same session, stimuli of five different intensities (P0 = undetected, P4 = very painful) were randomly presented (Boly et al., 2007). A specific involvement of left frontal cortex in top–down attentional control also agrees with recent EEG (Gomez, Marco-Pallares, & Grau, 2006) and fMRI studies (Hahn, Ross, & Stein, 2006; MacDonald, Cohen, Stenger, & Carter, 2000) showing that left dorsolateral prefrontal cortex (DLPFC) contributes to a system concerned with representing and maintaining attentional task demands. Specifically, a decrease in beta power over left inferior and middle frontal cortex was found during a period of directing attention to an upcoming visual stimulus (Gomez et al., 2006). Activation of a fronto-parietal network involved in top–down attentional control has frequently been reported in tasks of visual attention (Corbetta & Shulman, 2002; Kastner & Ungerleider, 2000). The leading role of frontal over sensory areas is supported also by a recent study investigating the temporal progression of prefrontal (DLPFC), frontal (frontal eye fields [FEFs]), and posterior parietal (lateral intraparietal area) cortical regions involved in top–down and bottom–up controlled visual attention (Buschman & Miller, 2007): By recording intracortical neuronal signals from multiple electrodes in monkeys, the authors found that the FEFs and DLPFC were activated 50 and 40 msec before a saccade, followed by an enhanced signal in parietal cortex not until 32 msec after the saccade for top–down visual spatial attention. This finding indicates that frontal cortex holds a leading role in top–down attentional control. The role of frontal cortex in attentional control has also been investigated in a recent EEG study recording theta band oscillations following a visuospatial cue (Green & McDonald, 2008). They found strong frontal activity in this frequency band related to the shifting of spatial attention, which was even preceded by parietal activity not found in any previous study. Transferred to the present masking paradigm, an attentional left frontal control system may prevent interference from task-irrelevant competing mask stimuli by maintaining the focus of attention on the target-receiving left hand. Whether the frontal desynchronization represents activation of DLPFC, the FEF, or both cannot be spatially disentangled here.

Centro-parietal Beta Synchronization

For single-stimulation trials as well as for double-stimulation trials with a short preceding ITI, we found a significant enhancement of the beta rhythmic activity at centro-parietal regions predicting conscious perception of target stimuli. The implications from these findings are twofold: First, non-Rolandic beta activity predicts somatosensory perception via synchronization in higher cortical regions. Second, the mechanism involved in single stimulus detection crucially differs from the mechanism in the masking condition with long preceding ITIs, but is comparable to the mechanism responsible for target detection with short ITIs.

The first issue is in line with findings by Linkenkaer-Hansen et al. (2004), who also reported a positive linear correlation between centro-parietal beta band power and detection rate of a near-threshold somatosensory stimulus. In the present study, while applying suprathreshold stimuli, yielding a detection rate of 75%, the single-stimulation condition resembles previous nonspatial studies applying near-threshold stimuli. In two recent MEG studies investigating visual motion detection (Donner et al., 2007) and an “attentional blink” paradigm (Gross et al., 2004), parietal beta amplitudes were found to be significantly larger before correct than incorrect behavioral choices and before hits than misses. These studies attribute central beta increase as a crucial gateway in a fronto-parietal attentional network. The fact that left frontal beta ERD is no longer significant in the topographic analysis for long ITIs is most probably due to the enhanced signal variance when reducing the data to one-fourth in this specific condition compared to the average ITI.

We assume that the difference between the results from the single left and short ITI double-stimulation conditions and the “long ITI double-stimulation condition” results from two different underlying mechanisms leading to conscious perception: Although a centro-parietal beta increase may index attentional network communication as a prerequisite for high performance, the left frontal and sensorimotor mu/beta decrease characterizes the strong spatial component of the current task, requiring top–down spatial-selective attention to resist against the strong mask stimulus applied at the other hand. We suggest that the duration of the preceding ITI modulates expectancy of the target stimulus (Romei et al., 2007), represented in preparation of spatially coordinating and stimulus processing areas. On the other hand, following a short ITI, the perceptual fate of the target stimulus could mainly depend on the general level of an attentive brain state. The idea that the expectancy of a target stimulus increases with increasing preceding ITI is further corroborated by the significantly higher target detection rate for long compared to short ITIs.

The Absolute Level of Mu/Beta Amplitudes Correlates with Resistance to Masking

The correlation analysis indicates that, for each subject, the success of resisting against a masking stimulus was nearly dependent on the individual level of prestimulus mu and beta band activity in sensorimotor as well as frontal cortex. Of note, for the correlation across subjects, the z-transformed absolute prestimulus amplitude values were taken. The implications from the correlation analysis are different from the topographic and time-course analyses, where we showed a “within-subjects across-trials” difference. The latter shows that, independent of the different levels of prestimulus mu/beta amplitudes, perception rate is related to the relative difference between the amplitudes for perceived and masked target stimuli. The correlation analysis, in addition, shows that also at the interindividual level, detection rate increases with decreasing absolute mu/beta amplitude size. This result is in line with the finding by Hanslmayr et al. (2007), who found significant lower occipital alpha power for the subjects in the “perceivers” group than for those in the “nonperceivers” group.

Prestimulus Rhythmic Activity and Poststimulus ERD/Evoked Potentials

Conscious perception of the left target stimulus is characterized by a strong ERD in the mu band in the contralateral hemisphere, and in the beta band in the contralateral hemisphere and in ipsilateral frontal areas emerging at 200–300 msec after stimulus onset. Of note, for double stimulation, a right foot response is conducted for perceived (left button press) as well as masked (right button press) trials. Thus, motor response-related activity and activity related to the mask stimulus is cancelled out and the results show the mere difference of the two perceptual states. The current results imply that not only somatosensory stimulus processing itself (Salenius et al., 1997; Pfurtscheller, 1981) but also whether the stimulus finds an entry into conscious perception determines the strength of suppression of the Rolandic rhythms. This is also in line with our previous finding that following double stimulation, conscious perception of a target stimulus is expressed in enhanced P100 and N140 SEP components (Schubert et al., 2006). Furthermore, these SEP effects were, as the present ERDs, most pronounced in the right parietal (P100) and bilateral frontal (N140) areas, probably representing the organization of the sensory input into spatial coordinates (Desmedt & Tomberg, 1989; Goldman-Rakic, 1987; Mountcastle, 1978), which have been tightly linked to conscious perception (Dehaene & Naccache, 2001).

Investigating visual spatial attention, Sauseng et al. (2005) could show that directed attention leads to alpha suppression in relevant brain areas. We suggest that prestimulus spatial attentional top–down influence induces enhanced excitability of somatosensory cortex. This enables entry into a conscious percept which is accompanied by a stronger activation/ERD of a fronto-parietal network.

At first glance, it may appear surprising that amplitude changes of prestimulus neuronal oscillatory activity in S1 can predict the detection of a somatosensory stimulus, although early (<100 msec) poststimulus evoked responses in S1 have been found unrelated to the conscious perception of such stimuli (Sarri et al., 2006; Schubert et al., 2006; Eimer et al., 2002). Yet, this dissociation is in line with previous MEG results showing that early evoked responses generated in S1 (N20m, P35m, and P60m) were stable in amplitude and latency even in face of a large variation in the prestimulus mu rhythm (Nikouline et al., 2000). Consequently, we propose that the initial processing of a suprathreshold stimulus in S1 is largely independent of the ongoing local network state (as indexed by the prestimulus level of ongoing mu/beta oscillations) and additive to it (Arieli, Sterkin, Grinvald, & Aertsen, 1996). Thus, obligatory early evoked components are generated for both perceived and unperceived stimuli. However, lower prestimulus mu/beta oscillations do predispose S1 for the perceptual detection of a target stimulus, which eventually is reflected by increased evoked components of longer peak latency (>100 msec; Schubert et al., 2006). This is in agreement with earlier findings in the visual system where low occipital prestimulus alpha power was predictive of larger stimulus evoked visual components at latencies longer than 100 msec (Rahn & Basar, 1993; Brandt & Jansen, 1991).

Conclusion

We propose the following model for prestimulus brain states that prevent masking: If the attentional control system of left frontal cortex becomes engaged prior to stimulus arrival, this is indexed by a desynchronization of regional beta rhythmic activity. It directs sensory cortex, which is expected to receive a task-critical target stimulus (here right primary somatosensory cortex) not to drift back into its idling state. This maintained desynchronization of the pericentral mu/beta rhythms then primes somatosensory cortex for the intensified processing of an upcoming target stimulus, leading to enhanced long-latency SEP components (>100 msec), which arise after the completion of obligatory early components that themselves are not predictive of conscious perception. This enhanced processing renders a target resistant against masking by a competing stimulus, and thus, allows for entry into conscious perception. As the targets are embedded into a stream of strong competing environmental stimuli, the present experimental paradigm models an ecologically valid condition.

Acknowledgments

This work was supported by the Bundesministerium für Bildung und Forschung (BMBF) (FKZ 16SV2234) and the Bernstein Center for Computational Neuroscience Berlin (project C4). We thank Steven Lemm for supporting the data analysis and Vadim Nikulin for discussions.

Reprint requests should be sent to Ruth Schubert, Neurophysics Group, Department of Neurology and Clinical Neurophysiology, Charité—University Medicine Berlin, Hindenburgdamm 30, 12200 Berlin, Germany, or via e-mail: ruth.schubert@charite.de.

REFERENCES

REFERENCES
Arieli
,
A.
,
Sterkin
,
A.
,
Grinvald
,
A.
, &
Aertsen
,
A.
(
1996
).
Dynamics of ongoing activity: Explanation of the large variability in evoked cortical responses.
Science
,
273
,
1868
1871
.
Babiloni
,
C.
,
Vecchio
,
F.
,
Bultrini
,
A.
,
Luca
,
R. G.
, &
Rossini
,
P. M.
(
2006
).
Pre- and poststimulus alpha rhythms are related to conscious visual perception: A high-resolution EEG study.
Cerebral Cortex
,
16
,
1690
1700
.
Bastiaansen
,
M. C.
, &
Brunia
,
C. H.
(
2001
).
Anticipatory attention: An event-related desynchronization approach.
International Journal of Psychophysiology
,
43
,
91
107
.
Boly
,
M.
,
Balteau
,
E.
,
Schnakers
,
C.
,
Degueldre
,
C.
,
Moonen
,
G.
,
Luxen
,
A.
,
et al
(
2007
).
Baseline brain activity fluctuations predict somatosensory perception in humans.
Proceedings of the National Academy of Sciences, U.S.A.
,
104
,
12187
12192
.
Brandt
,
M. E.
, &
Jansen
,
B. H.
(
1991
).
The relationship between prestimulus-alpha amplitude and visual evoked potential amplitude.
International Journal of Neuroscience
,
61
,
261
268
.
Buschman
,
T. J.
, &
Miller
,
E. K.
(
2007
).
Top–down versus bottom–up control of attention in the prefrontal and posterior parietal cortices.
Science
,
315
,
1860
1862
.
Corbetta
,
M.
, &
Shulman
,
G. L.
(
2002
).
Control of goal-directed and stimulus-driven attention in the brain.
Nature Reviews Neuroscience
,
3
,
201
215
.
de Lafuente
,
V.
, &
Romo
,
R.
(
2005
).
Neuronal correlates of subjective sensory experience.
Nature Neuroscience
,
8
,
1698
1703
.
Dehaene
,
S.
, &
Naccache
,
L.
(
2001
).
Towards a cognitive neuroscience of consciousness: Basic evidence and a workspace framework.
Cognition
,
79
,
1
37
.
Desmedt
,
J. E.
, &
Tomberg
,
C.
(
1989
).
Mapping early somatosensory evoked potentials in selective attention: Critical evaluation of control conditions used for titrating by difference the cognitive P30, P40, P100 and N140.
Electroencephalography and Clinical Neurophysiology
,
74
,
321
346
.
Donner
,
T. H.
,
Siegel
,
M.
,
Oostenveld
,
R.
,
Fries
,
P.
,
Bauer
,
M.
, &
Engel
,
A. K.
(
2007
).
Population activity in the human dorsal pathway predicts the accuracy of visual motion detection.
Journal of Neurophysiology
,
98
,
345
359
.
Eimer
,
M.
, &
Forster
,
B.
(
2003
).
Modulations of early somatosensory ERP components by transient and sustained spatial attention.
Experimental Brain Research
,
151
,
24
31
.
Eimer
,
M.
,
Forster
,
B.
, &
van Velzen
,
J.
(
2003
).
Anterior and posterior attentional control systems use different spatial reference frames: ERP evidence from covert tactile-spatial orienting.
Psychophysiology
,
40
,
924
933
.
Eimer
,
M.
,
Maravita
,
A.
,
Van Velzen
,
J.
,
Husain
,
M.
, &
Driver
,
J.
(
2002
).
The electrophysiology of tactile extinction: ERP correlates of unconscious somatosensory processing.
Neuropsychologia
,
40
,
2438
2447
.
Engel
,
A. K.
,
Fries
,
P.
, &
Singer
,
W.
(
2001
).
Dynamic predictions: Oscillations and synchrony in top–down processing.
Nature Reviews Neuroscience
,
2
,
704
716
.
Ergenoglu
,
T.
,
Demiralp
,
T.
,
Bayraktaroglu
,
Z.
,
Ergen
,
M.
,
Beydagi
,
H.
, &
Uresin
,
Y.
(
2004
).
Alpha rhythm of the EEG modulates visual detection performance in humans.
Brain Research, Cognitive Brain Research
,
20
,
376
383
.
Forster
,
B.
, &
Eimer
,
M.
(
2004
).
The attentional selection of spatial and non-spatial attributes in touch: ERP evidence for parallel and independent processes.
Biological Psychology
,
66
,
1
20
.
Fries
,
P.
,
Reynolds
,
J. H.
,
Rorie
,
A. E.
, &
Desimone
,
R.
(
2001
).
Modulation of oscillatory neuronal synchronization by selective visual attention.
Science
,
291
,
1560
1563
.
Goldman-Rakic
,
P. S.
(
1987
).
Circuitry of primate prefrontal cortex and regulation of behavior by representational memory.
In F. Plum (Ed.),
Handbook of physiology. Vol. 5: The nervous system
(pp.
373
417
).
Bethesda, MD
:
American Physiological Society
.
Gomez
,
C. M.
,
Marco-Pallares
,
J.
, &
Grau
,
C.
(
2006
).
Location of brain rhythms and their modulation by preparatory attention estimated by current density.
Brain Research
,
1107
,
151
160
.
Green
,
J. J.
, &
McDonald
,
J. J.
(
2008
).
Electrical neuroimaging reveals timing of attentional control activity in human brain.
PLoS Biology
,
6
,
730
738
.
Gross
,
J.
,
Schmitz
,
F.
,
Schnitzler
,
I.
,
Kessler
,
K.
,
Shapiro
,
K.
,
Hommel
,
B.
,
et al
(
2004
).
Modulation of long-range neural synchrony reflects temporal limitations of visual attention in humans.
Proceedings of the National Academy of Sciences, U.S.A.
,
101
,
13050
13055
.
Hahn
,
B.
,
Ross
,
T. J.
, &
Stein
,
E. A.
(
2006
).
Neuroanatomical dissociation between bottom–up and top–down processes of visuospatial selective attention.
Neuroimage
,
32
,
842
853
.
Hanslmayr
,
S.
,
Aslan
,
A.
,
Staudigl
,
T.
,
Klimesch
,
W.
,
Herrmann
,
C. S.
, &
Bauml
,
K. H.
(
2007
).
Prestimulus oscillations predict visual perception performance between and within subjects.
Neuroimage
,
37
,
1465
1473
.
Jones
,
S. R.
,
Pritchett
,
D. L.
,
Stufflebeam
,
S. M.
,
Hämäläinen
,
M.
, &
Moore
,
C. I.
(
2007
).
Neural correlates of tactile detection: A combined magnetoencephalography and biophysically based computational modeling study.
Journal of Neuroscience
,
27
,
10751
10764
.
Josiassen
,
R. C.
,
Shagass
,
C.
,
Roemer
,
R. A.
,
Slepner
,
S.
, &
Czartorysky
,
B.
(
1990
).
Early cognitive components of somatosensory event-related potentials.
International Journal of Psychophysiology
,
9
,
139
149
.
Kastner
,
S.
, &
Ungerleider
,
L. G.
(
2000
).
Mechanisms of visual attention in the human cortex.
Annual Review of Neuroscience
,
23
,
315
341
.
Klimesch
,
W.
(
1999
).
EEG alpha and theta oscillations reflect cognitive and memory performance: A review and analysis.
Brain Research, Brain Research Reviews
,
29
,
169
195
.
Klimesch
,
W.
,
Sauseng
,
P.
, &
Gerloff
,
C.
(
2003
).
Enhancing cognitive performance with repetitive transcranial magnetic stimulation at human individual alpha frequency.
European Journal of Neuroscience
,
17
,
1129
1133
.
Klimesch
,
W.
,
Sauseng
,
P.
, &
Hanslmayr
,
S.
(
2007
).
EEG alpha oscillations: The inhibition-timing hypothesis.
Brain Research Reviews
,
53
,
63
88
.
Libet
,
B.
,
Alberts
,
W. W.
,
Wright
,
E. W.
, Jr., &
Feinstein
,
B.
(
1967
).
Responses of human somatosensory cortex to stimuli below threshold for conscious sensation.
Science
,
158
,
1597
1600
.
Linkenkaer-Hansen
,
K.
,
Nikulin
,
V. V.
,
Palva
,
S.
,
Ilmoniemi
,
R. J.
, &
Palva
,
J. M.
(
2004
).
Prestimulus oscillations enhance psychophysical performance in humans.
Journal of Neuroscience
,
24
,
10186
10190
.
MacDonald
,
A. W.
, III,
Cohen
,
J. D.
,
Stenger
,
V. A.
, &
Carter
,
C. S.
(
2000
).
Dissociating the role of the dorsolateral prefrontal and anterior cingulate cortex in cognitive control.
Science
,
288
,
1835
1838
.
Meador
,
K. J.
,
Ray
,
P. G.
,
Day
,
L.
,
Ghelani
,
H.
, &
Loring
,
D. W.
(
1998
).
Physiology of somatosensory perception: Cerebral lateralization and extinction.
Neurology
,
51
,
721
727
.
Michie
,
P. T.
,
Bearpark
,
H. M.
,
Crawford
,
J. M.
, &
Glue
,
L. C.
(
1987
).
The effects of spatial selective attention on the somatosensory event-related potential.
Psychophysiology
,
24
,
449
463
.
Mountcastle
,
V. B.
(
1978
).
Brain mechanisms for directed attention.
Journal of the Royal Society of Medicine
,
71
,
14
28
.
Nikouline
,
V. V.
,
Wikström
,
H.
,
Linkenkaer-Hansen
,
K.
,
Kesäniemi
,
M.
,
Ilmoniemi
,
R. J.
, &
Huttunen
,
J.
(
2000
).
Somatosensory evoked magnetic fields: Relation to pre-stimulus mu rhythm.
Clinical Neurophysiology
,
111
,
1227
1233
.
Palva
,
S.
,
Linkenkaer-Hansen
,
K.
,
Näätänen
,
R.
, &
Palva
,
J. M.
(
2005
).
Early neural correlates of conscious somatosensory perception.
Journal of Neuroscience
,
25
,
5248
5258
.
Pfurtscheller
,
G.
(
1981
).
Central beta rhythm during sensorimotor activities in man.
Electroencephalography and Clinical Neurophysiology
,
51
,
253
264
.
Pfurtscheller
,
G.
, &
Lopes da Silva
,
F. H.
(
1999
).
Event-related EEG/MEG synchronization and desynchronization: Basic principles.
Clinical Neurophysiology
,
110
,
1842
1857
.
Rahn
,
E.
, &
Basar
,
E.
(
1993
).
Enhancement of visual evoked potentials by stimulation during low prestimulus EEG stages.
International Journal of Neuroscience
,
72
,
123
136
.
Ray
,
P. G.
,
Meador
,
K. J.
,
Smith
,
J. R.
,
Wheless
,
J. W.
,
Sittenfeld
,
M.
, &
Clifton
,
G. L.
(
1999
).
Physiology of perception: Cortical stimulation and recording in humans.
Neurology
,
52
,
1044
1049
.
Romei
,
V.
,
Brodbeck
,
V.
,
Michel
,
C.
,
Amedi
,
A.
,
Pascual-Leone
,
A.
, &
Thut
,
G.
(
2007
).
Spontaneous fluctuations in posterior {alpha}-band EEG activity reflect variability in excitability of human visual areas.
Cerebral Cortex
,
18
,
2010
2018
.
Romei
,
V.
,
Rihs
,
T.
,
Brodbeck
,
V.
, &
Thut
,
G.
(
2008
).
Resting electroencephalogram alpha-power over posterior sites indexes baseline visual cortex excitability.
NeuroReport
,
19
,
203
208
.
Salenius
,
S.
,
Schnitzler
,
A.
,
Salmelin
,
R.
,
Jousmäki
,
V.
, &
Hari
,
R.
(
1997
).
Modulation of human cortical rolandic rhythms during natural sensorimotor tasks.
Neuroimage
,
5
,
221
228
.
Salmelin
,
R.
, &
Hari
,
R.
(
1994
).
Spatiotemporal characteristics of sensorimotor neuromagnetic rhythms related to thumb movement.
Neuroscience
,
60
,
537
550
.
Sarri
,
M.
,
Blankenburg
,
F.
, &
Driver
,
J.
(
2006
).
Neural correlates of crossmodal visual–tactile extinction and of tactile awareness revealed by fMRI in a right-hemisphere stroke patient.
Neuropsychologia
,
44
,
2398
2410
.
Sauseng
,
P.
,
Klimesch
,
W.
,
Stadler
,
W.
,
Schabus
,
M.
,
Doppelmayr
,
M.
,
Hanslmayr
,
S.
,
et al
(
2005
).
A shift of visual spatial attention is selectively associated with human EEG alpha activity.
European Journal of Neuroscience
,
22
,
2917
2926
.
Schubert
,
R.
,
Blankenburg
,
F.
,
Lemm
,
S.
,
Villringer
,
A.
, &
Curio
,
G.
(
2006
).
Now you feel it–Now you don't: ERP correlates of somatosensory awareness.
Psychophysiology
,
43
,
31
40
.
Steriade
,
M.
, &
Llinas
,
R. R.
(
1988
).
The functional states of the thalamus and the associated neuronal interplay.
Physiological Reviews
,
68
,
649
742
.
Thut
,
G.
,
Nietzel
,
A.
,
Brandt
,
S. A.
, &
Pascual-Leone
,
A.
(
2006
).
Alpha-band electroencephalographic activity over occipital cortex indexes visuospatial attention bias and predicts visual target detection.
Journal of Neuroscience
,
26
,
9494
9502
.
Tiihonen
,
J.
,
Hari
,
R.
, &
Hämäläinen
,
M.
(
1989
).
Early deflections of cerebral magnetic responses to median nerve stimulation.
Electroencephalography and Clinical Neurophysiology
,
74
,
290
296
.
van Dijk
,
H.
,
Schoffelen
,
J.-M.
,
Oostenveld
,
R.
, &
Jensen
,
O.
(
2008
).
Prestimulus oscillatory activity in the alpha band predicts visual discrimination ability [abstract].
Journal of Neuroscience
,
28
,
1816
1823
.
Worden
,
M. S.
,
Foxe
,
J. J.
,
Wang
,
N.
, &
Simpson
,
G. V.
(
2000
).
Anticipatory biasing of visuospatial attention indexed by retinotopically specific alpha-band electroencephalography increases over occipital cortex.
Journal of Neuroscience
,
20
,
RC63
.
Yamagishi
,
N.
,
Callan
,
D. E.
,
Anderson
,
S. J.
, &
Kawato
,
M.
(
2008
).
Attentional changes in pre-stimulus oscillatory activity within early visual cortex are predictive of human visual performance.
Brain Research
,
1197
,
115
122
.
Zhu
,
Z.
,
Disbrow
,
E. A.
,
Zumer
,
J. M.
,
McGonigle
,
D. J.
, &
Nagarajan
,
S. S.
(
2007
).
Spatiotemporal integration of tactile information in human somatosensory cortex.
BMC Neuroscience
,
8
,
21
.