Elsevier

Neuropsychologia

Volume 96, February 2017, Pages 9-18
Neuropsychologia

Pre-stimulus alpha oscillations over somatosensory cortex predict tactile misperceptions

https://doi.org/10.1016/j.neuropsychologia.2016.12.030Get rights and content

Highlights

  • Pre-stimulus alpha brain activity influences detection of peri-threshold touch.

  • Alpha power over somatosensory cortex was measured using EEG to hits and false alarms.

  • The probability of reporting touch decreased as contralateral alpha power increased.

  • Ipsilateral alpha power showed a similar relationship for false alarms.

  • Accounts of the role of pre-stimulus power need to consider true and false perceptions.

Abstract

Fluctuations of pre-stimulus oscillatory activity in the somatosensory alpha band (8–14 Hz) observed using human EEG and MEG have been shown to influence the detection of supra- and peri-threshold somatosensory stimuli. However, some reports of touch occur even without a stimulus. We investigated the possibility that pre-stimulus alpha oscillations might also influence these false reports of touch – known as tactile misperceptions. We recorded EEG while participants performed the Somatic Signal Detection Task (SSDT), in which participants must detect brief, peri-threshold somatosensory targets. We found that pre-stimulus oscillatory power in the somatosensory alpha range exhibited a negative linear relationship with reporting of touch at electrode clusters over both contralateral and ipsilateral somatosensory regions. As pre-stimulus alpha power increased, the probability of reporting a touch declined; as it decreased, the probability of reporting a touch increased. This relationship was stronger on trials without a somatosensory stimulus than on trials with a somatosensory stimulus, although was present for both trial types. Spatio-temporal cluster-based permutation analysis also found that pre-stimulus alpha was lower on trials when touch was reported – irrespective of whether it was present – over contralateral and ipsilateral somatosensory cortices, as well as left frontocentral areas. We argue that alpha power may reflect changes in response criterion rather than sensitivity alone. Low alpha power relates to a low barrier to reporting a touch even when one is not present, while high alpha power is linked to less frequent reporting of touch overall.

Introduction

Fluctuations of oscillatory activity in the somatosensory alpha range (8–14 Hz) observed using the human EEG and MEG have been shown to influence detection of supra-, peri-, and sub-threshold somatosensory targets. Specifically, detection of targets improves and declines according to the power and phase of ongoing alpha oscillations over (typically) contralateral primary somatosensory cortex (Ai and Ro, 2014, Jones et al., 2010, Linkenkaer-Hansen et al., 2004, Weisz et al., 2014, Zhang and Ding, 2010). Alpha has been linked to cortical excitability, which may be the underlying mechanism that links observed alpha with performance (e.g. Rajagovindan and Ding, 2010; Romei et al., 2008a; Romei et al., 2008b). However, the influence of oscillatory power on the false alarm rate – reports of stimuli which are not there – has often been neglected. In vision, some reports have suggested that alpha power may determine responses to absent stimuli (Limbach and Corballis, 2016) or misperceptions of one stimulus as two (Lange et al.,2013). Understanding how ongoing activity relates to both hits and false alarms may be critical to understanding the functional role of alpha oscillations in tactile perception.

Alpha power has been strongly linked to cortical excitability and inhibition (Jensen and Mazaheri, 2010, Klimesch et al., 2007, Romei et al., 2008a), with low alpha power indicating that a region is in a state of high excitability, and high alpha power indicating low excitability or inhibition. Trial-to-trial fluctuations in power might account for variations in the detectability of stimuli: A weak stimulus might be sufficient to produce a response from a region in a low-alpha, high-excitability state, but insufficient when that region is in a high-alpha, low-excitability state. Alpha oscillations may thus reflect pulses of cortical inhibition (Mathewson et al., 2011), as a sensory region adjusts to temporal expectations of the delivery of a stimulus by ensuring it is in a high-excitability state when a stimulus is expected. Top-down manipulation of alpha power may also be a mechanism for the suppression of irrelevant stimuli (Haegens et al., 2011), through the inhibition of cortical regions dealing with, for example, unattended regions of space. This may also be used to gate communication between different cortical regions (Jensen et al., 2012, Jensen and Mazaheri, 2010), allowing or preventing early sensory responses to reach consciousness (Weisz et al., 2014).

Although alpha has most frequently been shown to have a negative linear relationship with detection of visual targets, with lower alpha power increasing hit-rates (Busch et al., 2009, Ergenoglu et al., 2004, Hanslmayr et al., 2007), the pattern of the relationship between somatosensory alpha and performance is much more mixed. Several groups have reported a similar negative linear relationship between alpha and performance as that seen in vision, with low alpha power associated with better performance on both tactile detection tasks and simultaneity judgement tasks (Baumgarten et al., 2016, Jones et al., 2010, van Ede et al., 2012). This suggests that pre-stimulus somatosensory and visual alpha may share common functional roles and mechanisms. However, several other groups have reported a quadratic relationship between alpha and performance, with intermediate power levels associated with improved hit rates (Ai and Ro, 2014, Linkenkaer-Hansen et al., 2004, Weisz et al., 2014, Zhang and Ding, 2010).

Notably, the role of pre-stimulus alpha in false reports of a tactile or visual stimulus was not directly addressed in any of these studies. High cortical excitability may also be expected to increase false alarms if the state of the system is such that even a weak stimulus is sufficient to generate a large response. Participants may be more likely to erroneously respond to non-target stimuli or endogenously generated sensations as well as genuine stimuli. Consistent with this suggestion, a recent report suggests that alpha power during visual tasks may modify response criterion rather than visual sensitivity (Limbach and Corballis, 2016). They found that participants were more likely to report a stimulus when alpha was low than when alpha was high. Both hit rates and false alarms decreased as alpha power increased, suggesting a tightening of response criterion but no increase in sensitivity with increasing power. Similarly, Samaha et al. (2016) reported that alpha power negatively correlated with participants’ confidence in their responses in a visual two-choice orientation discrimination task, but did not predict the accuracy of those responses. Sherman et al. (2016) also report that alpha phase, which may index moment-by-moment excitability, determined response criterion in a visual detection task. Participants’ reports of the presence or absence of a stimulus were partially predicted by phase at stimulus onset, regardless of whether their decision was correct.

Although there is evidence suggesting that alpha oscillations may shift response criterion or confidence in visual tasks, evidence for shifting the response criterion in touch is, in contrast, lacking. In macaques, Haegens et al. (2014) reported that alpha oscillations in contralateral primary somatosensory cortex shifted response criterion rather than tactile sensitivity. Higher alpha power was associated with a lower criterion, and thus an increase in stimulus-present responses regardless of whether the stimulus was present. Notably, this is the opposite pattern to that reported by Limbach and Corballis (2016) in humans in vision, suggesting that somatosensory related alpha power may function in a different way than in alpha power visual tasks.

We examined alpha oscillations using electroencephalography (EEG) recorded while participants performed the Somatic Signal Detection Task (SSDT; Lloyd et al., 2008). In the SSDT, participants detect brief somatosensory stimuli delivered to their non-dominant hand at their individual perceptual threshold. In addition, an LED placed close to the non-dominant hand flashes during the SSDT, sometimes simultaneous with the tactile stimulus and sometimes on its own. The presence of the LED has previously been shown to increase false alarm rates (Brown et al., 2010, Lloyd et al., 2008, Lloyd et al., 2011). A common impediment to examining the relationship between alpha power and false alarms is the relative lack of false alarm trials to examine. False alarm rates in many tasks are typically below 15% (e.g. Busch et al., 2009; Mathewson et al., 2009). The SSDT typically yields false alarm rates of 15–20%. Thus, the SSDT offers an excellent paradigm for examining somatosensory performance and specifically for producing sufficient false alarms to quantitatively investigate how alpha oscillations relate to both true and false reports of touch.

Section snippets

Participants

In total, we recruited 32 right-handed participants (age range =18 to 45 years, mean age =22 years, S.D. =6 years, 28 female) drawn from the participant pool at the University of Leeds. Participants provided written, informed consent to take part and the experimental procedure was approved by the Ethical Committee of the School of Psychology at the University of Leeds. All participants had normal or corrected-to-normal vision and reported no tactile deficits.

Apparatus

The stimulus array comprised a

Behavioural data

Participants reported far more touches when touch was present than when it was absent [F(1,31)=66.9, p<.001, ƞ2g =.53], and more touches when light was present than when it was not [F(1,31)=35.69, p<.001, ƞ2g =.11]. In addition, there was a significant interaction between light and touch [F(1,31)=18.51, p<.001, ƞ2g=.02], (see Fig. 2). Post-hoc tests indicated that all means were significantly different from each other (all ps<.002). The interaction was driven by the noticeably larger effect of

Discussion

We examined the relationship between somatosensory alpha and the reporting of tactile sensations during the SSDT. We found that pre-stimulus alpha generally followed a negative linear relationship with reporting of touch, with higher alpha associated with a decrease in reports of touch, and lower alpha associated with an increase in reports of touch. The relationship between alpha power and reporting of touch interacted with the actual presence of touch in both hemispheres, exhibiting a weaker

Acknowledgements

This study was funded by the Biotechnology and Biological Sciences Research Council (BBSRC; BB/L006618/1). WeD received funding from CONICYT, Chile, Basal project FB0008. All authors declare no conflicts of interest.

References (50)

  • E. Maris et al.

    Nonparametric statistical testing of EEG- and MEG-data

    J. Neurosci. Methods

    (2007)
  • H. Nolan et al.

    FASTER: fully automated statistical thresholding for EEG artifact rejection

    J. Neurosci. Methods

    (2010)
  • M. Ploner et al.

    Oscillatory activity reflects the excitability of the human somatosensory system

    NeuroImage

    (2006)
  • V. Romei et al.

    Sounds reset rhythms of visual cortex and corresponding human visual perception

    Curr. Biol.

    (2012)
  • J. Vosskuhl et al.

    BOLD signal effects of transcranial alternating current stimulation (tACS) in the alpha range: a concurrent tACS–fMRI study

    NeuroImage

    (2016)
  • L. Ai et al.

    The phase of prestimulus alpha oscillations affects tactile perception

    J. Neurophysiol.

    (2014)
  • D. Bates et al.

    Fitting linear mixed-effects models using lme4

    J. Stat. Softw.

    (2015)
  • Bates, D., Mächler, M., Bolker, B., Walker, S., 2016. lme4: Linear mixed-effects models using Eigen and S4. (Version...
  • T.J. Baumgarten et al.

    Prestimulus alpha power influences tactile temporal perceptual discrimination and confidence in decisions

    Cereb. Cortex

    (2016)
  • N.A. Busch et al.

    The phase of ongoing EEG oscillations predicts visual perception

    J. Neurosci.

    (2009)
  • M.J. Gray et al.

    Oscillatory recruitment of bilateral visual cortex during spatial attention to competing rhythmic inputs

    J. Neurosci.

    (2015)
  • S. Haegens et al.

    Top-down controlled alpha band activity in somatosensory areas determines behavioral performance in a discrimination task

    J. Neurosci.

    (2011)
  • S. Haegens et al.

    Somatosensory anticipatory alpha activity increases to suppress distracting input

    J. Cogn. Neurosci.

    (2012)
  • S. Haegens et al.

    Thalamocortical rhythms during a vibrotactile detection task

    Proc. Natl. Acad. Sci. U.S.A.

    (2014)
  • C.S. Herrmann et al.

    Shaping intrinsic neural oscillations with periodic stimulation

    J. Neurosci.

    (2016)
  • Cited by (0)

    View full text