On the relationship between cortical excitability and visual oscillatory responses — A concurrent tDCS–MEG study

Highlights

• We stimulated participants' visual cortices with tDCS during MEG recording.

• Expected visual oscillatory activity could be recovered during tDCS stimulation.

• tDCS did not alter oscillatory activity in the alpha- or gamma-frequency bands.

Abstract

Neuronal oscillations in the alpha band (8–12 Hz) in visual cortex are considered to instantiate ‘attentional gating’ via the inhibition of activity in regions representing task-irrelevant parts of space. In contrast, visual gamma-band activity (40–100 Hz) is regarded as representing a bottom-up drive from incoming visual information, with increased synchronisation producing a stronger feedforward impulse for relevant information. However, little is known about the direct relationship between excitability of the visual cortex and these oscillatory mechanisms. In this study we used transcranial direct current stimulation (tDCS) in an Oz–Cz montage in order to stimulate visual cortex, concurrently recording whole-brain oscillatory activity using magnetoencephalography (MEG) whilst participants performed a visual task known to produce strong modulations of alpha- and gamma-band activity. We found that visual stimuli produced expected modulations of alpha and gamma – presenting a moving annulus stimulus led to a strong gamma increase and alpha decrease – and that this pattern was observable both during active (anodal and cathodal) tDCS and sham tDCS. However, tDCS did not seem to produce systematic alterations of these oscillatory responses. The present study thus demonstrates that concurrent tDCS/MEG of the visual system is a feasible tool for investigating visual neuronal oscillations, and we discuss potential reasons for the apparent inability of tDCS to effectively change the amplitude of visual stimulus induced oscillatory responses in the current study.

Introduction

A substantial body of evidence suggests that alpha and gamma oscillations are signatures of fundamental neural mechanisms of visual stimulus processing. Oscillatory activity in the alpha band is suggested to represent active inhibition of task-irrelevant brain regions by means of neuronal gating (Klimesch et al., 2007, Jensen and Mazaheri, 2010). Commensurate with this, alpha oscillations increase in regions of sensory cortex anticipating distracting information (Worden et al., 2000, Thut et al., 2006, Haegens et al., 2011, Van Ede et al., 2011) in a manner that predicts behaviour (Handel et al., 2010). In contrast, gamma-band activity is thought to represent active processing of stimuli, indexing enhanced feedforward propagation of stimulus representations via increased neuronal gain (Salinas and Sejnowski, 2001, Tiesinga et al., 2004), which may be a fundamental mechanism for computation across regions (Fries et al., 2007, Fries, 2009). Correspondingly, gamma-band activity in sensory cortex increases with attention (Müller et al., 2000, Fries et al., 2001, Ray et al., 2008, Koelewijn et al., 2013) and this increase predicts enhanced behavioural judgements (Hoogenboom et al., 2006). Gamma power and frequency are also dependent on several low-level stimulus features, including visual contrast (Henrie and Shapley, 2005, Schadow et al., 2007, Ray and Maunsell, 2010), eccentricity (Busch et al., 2004, Van Pelt and Fries, 2013), noise level, spatial frequency (Fründ et al., 2007), and size (Jia et al., 2013), as well as differing for gratings and plaids (Lima et al., 2010).

Although the ‘gating by inhibition’ framework (Jensen and Mazaheri, 2010) posits that alpha activity serves an inhibitory role and gamma represents bottom-up excitatory drive, little is currently known about the direct relationship between cortical excitability and these oscillatory measures. Exploring this relationship necessitates exogenous manipulation of excitability. This can be achieved pharmacologically (Bauer et al., 2012, Lozano-Soldevilla et al., 2014) however drug manipulations are extremely non-focal, meaning that specific conclusions cannot be drawn about the excitability of a given cortical region. Patterned transcranial magnetic stimulation (TMS) has also been shown to reversibly inhibit or excite a cortical region for a period of up to one hour following stimulation (Huang et al., 2005), and – in separate studies – has also been shown to impact both alpha and gamma oscillations during attentional task performance (Sauseng et al., 2011, Marshall et al., 2015). However, TMS can only be used to induce offline changes – meaning, changes that follow a period of stimulation – in cortical excitability resulting from synaptic changes, rather than online changes in membrane polarisation during stimulation.

A recent study (Soekadar et al., 2013) demonstrated for the first time that transcranial direct current stimulation (tDCS) can be applied during recording of electrophysiological brain activity with magnetoencephalography (MEG). This novel approach permits whole-brain measurement of neuronal oscillations concurrently with reversible, manipulation of cortical excitability during stimulation (i.e., an online effect). tDCS involves the application of weak, direct electric currents via the placement of electrodes on the scalp; current flows from the anode to the cathode and some portion of this current flows through the cortex. tDCS has been shown to produce both online (during stimulation) and offline (after stimulation) effects on cortical excitability; invasive recordings in animals have demonstrated that positive polarisation increases neuronal firing, whereas negative polarisation decreases it (Bindman et al., 1964, Fröhlich and McCormick, 2010). tDCS studies in humans have largely targeted the motor cortex (Horvath et al., 2014), however several studies have also shown effects of visual cortical stimulation on a variety of behavioural output measures (Antal et al., 2004b, Accornero et al., 2007, Medina et al., 2013, Peters et al., 2013, Makovski and Lavidor, 2014). This makes tDCS an extremely useful technique for exploring the relationship between cortical excitability and visual alpha and gamma oscillations.

To that end we combined tDCS of the visual cortex with concurrent MEG. We adapted a behavioural task known to elicit robust modulations of visual alpha and gamma oscillations (Hoogenboom et al., 2006), and applied short blocks of anodal, cathodal, and sham tDCS whilst participants performed a visual task in the MEG. We chose to use short two-minute tDCS blocks to avoid the build-up of offline effects carrying over into subsequent blocks, as previous studies indicate that a minimum of 3 min of stimulation is required for effects to persist after tDCS is discontinued (Nitsche and Paulus, 2000). Since gamma is proposed to indicate a feedforward drive, whereas alpha indicates an inhibitory state, we hypothesized that increasing visual cortical excitability with anodal tDCS would increase the amplitude of the gamma-band response and decrease the alpha-band response to the visual stimulus, and that cathodal tDCS would have the opposite effect; reducing visual cortical excitability and thus decreasing gamma power and increasing alpha power.