Motor and attentional mechanisms involved in social interaction—Evidence from mu and alpha EEG suppression

NeuroImage 58 (2011) 895–904

Contents lists available at ScienceDirect

NeuroImage j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / y n i m g

Motor and attentional mechanisms involved in social interaction—Evidence from mu and alpha EEG suppression Anat Perry a,⁎, Libi Stein b, Shlomo Bentin a, c a b c

Department of Psychology, Hebrew University, Jerusalem, Israel Department of Cognitive Sciences, Hebrew University, Jerusalem, Israel The Interdisciplinary Center for Neural Computation Hebrew University, Jerusalem, Israel

a r t i c l e

i n f o

Article history: Received 22 March 2011 Revised 20 June 2011 Accepted 21 June 2011 Available online 30 June 2011 Keywords: Alpha rhythms Mu rhythms EEG Social interaction

a b s t r a c t Mu rhythms are EEG oscillations in the 8–13 Hz recorded at sites located roughly over the sensory-motor cortex. There is reliable evidence that the amplitude of mu rhythms is reduced when the participant performs a motor act (mu suppression). Recent studies found mu suppression not only in response to actual movements but also while the participant observes actions executed by someone else. This finding putatively associates the mu suppression to the activity of a mirror neurons system which, in humans, has been suggested to contribute to social skills. In the present study we explored the effects of different levels of social interaction on mu suppression. Participants observed dynamic displays of hand gestures performing actions used in the Rock–Scissors–Paper game. In different blocks, participants passively viewed identical video clips with no game context and in the context of a game, or while being actually engaged in the game either by imagining actions or by actual playing. As a baseline for calculating mu suppression we used a dynamic display of a rolling ball. In addition, to isolate the social aspect of the actual movements, participants performed the same acts outside the game context. Mu suppression was larger while participants were engaged in the social game than when they passively looked at the “opponent” actions or when they performed movements without the game context. This effect was found while viewing the opponent play as well as while actually playing, which supports the view that mu suppression is affected not only by motion, but also by the social context of the motion. However, we did not find differences in mu suppression between perception segments in which the participant did not actually play. Furthermore, in all perception segments occipital alpha suppression was more robust than mu suppression suggesting the involvement of a strong attentional component. While actually playing, however, mu suppression was stronger than alpha suppression. © 2011 Elsevier Inc. All rights reserved.

Introduction As human beings, we are usually aware of our own beliefs, intents, desires, pretence, knowledge, etc., and are able to understand that others also have beliefs, desires and intentions that may be similar or different from ours. Numerous studies have investigated the neural substrates of human's social skills (For reviews, see Adolphs, 2003; Lieberman, 2007; Pfeifer et al. 2007) but the neural mechanisms that enables humans to gain such knowledge and the nature of the information that feeds into it are still under debate.

Abbreviations: ST, Simulation Theory; EEG, Electroencephalography; MEG, Magnetoencephalography; fMRI, functional Magnetic Resonance Imaging; TMS, Transcranial Magnetic Stimulation; MNS, Mirror Neuron System; hMNS, human Mirror Neuron System; FFT, Fast Fourier Transform. ⁎ Corresponding author at: Department of Psychology, Hebrew University, Jerusalem, 91905 Israel. Fax: +972 25825659. E-mail addresses: [email protected] (A. Perry), [email protected] (L. Stein), [email protected] (S. Bentin). 1053-8119/$ – see front matter © 2011 Elsevier Inc. All rights reserved. doi:10.1016/j.neuroimage.2011.06.060

A prevailing view, which stems from the social–cognitive Simulation Theory (ST; Carruthers, 1996; Davis & Stone, 1995), is that the information which at least some of our abilities to understand others’ state of mind relies on, is not only sensory but includes also a significant motor component (e.g. Jackson & Decety, 2004; Rizzolatti et al., 2001). Further, ST purports that the general mechanism whereby humans are able to generate knowledge of other minds involves simulating the actions performed by others and associating the simulated action with motor representations of our own internal states, motivations, and intentions (Agnew et al., 2007; Keysers & Perrett, 2004; Niedenthal et al., 2005). This idea fits with the video motor framework of action (Prinz, 2005), which assumes a common representational format for perception and action. Whereas there are alternative theories which attempt to account for our abilities to “get into the others’ minds”, the biological feasibility of ST is supported by the seminal discovery of mirror neurons in the monkey. Mirror neurons are a particular class of visuo-motor neurons that discharge both when the monkey does a particular goal-directed action and when it observes another individual (monkey or human) doing a

896

A. Perry et al. / NeuroImage 58 (2011) 895–904

similar action (di Pellegrino et al., 1992; Gallese et al., 1996; Rizzolatti et al., 1996). The neurophysiological properties of the mirror neurons and their putatively important role in cognition raised the question of whether similar motor-resonance mechanisms exists in humans and, assuming that they do, what role such networks play in social cognition. In the last decade, an allegedly human Mirror Neurons System (hMNS) has been extensively investigated through hemodynamic neuroimaging (e.g. Buccino et al., 2001; Decety et al., 2002; Grezes et al., 2003; for reviews see Iacoboni & Dapretto, 2006 and Rizzolatti & Sinigaglia, 2010) and many studies suggested that a mirror-like system might, indeed, contribute to the understanding of social behavior (Carr et al., 2003; Happe et al., 1996; Leslie et al., 2004; Singer et al., 2004; Wicker et al., 2003; for recent reviews see Carrington & Bailey, 2009; Hari & Kujala, 2009; Iacoboni, 2009). Additional evidence for a hMNS has also been found using electrophysiological measures such as modulation of motor evoked potentials elicited by TMS (Fadiga et al., 1995), intracranial single-unit recordings (Mukamel et al., 2010), magnetoencephalography (MEG; e.g. Hari et al., 1997) and electroencephalography (EEG). The latter line of research focused particularly on the modulation of mu rhythms, which are EEG oscillations within the range of 8–13 Hz measured over sensory-motor regions. Mu rhythms are desynchronized and their power attenuated when engaging in motor activity (Gastaut, 1952), and also while observing actions executed by someone else (e.g. Gastaut, 1952; Muthukumaraswamy et al., 2004) or even imagining performing an action (Pfurtscheller et al., 2008). These characteristics led authors to tentatively link the suppression of mu rhythms with the hMNS (for a review, see Pineda, 2005). In the last few years, several studies of typical participants linked EEG mu suppression to higher social information processing including Theory of Mind (Pineda & Hecht, 2009) and perception of others' pain (Cheng et al., 2008; Perry et al., 2010a). Supporting a link between social cognition and this EEG manifestation, several studies of autistic spectrum disorders (ASD) found abnormal mu suppression when ASD individuals viewed actions performed by others despite normal suppression while performing the same actions (Martineau et al., 2004; Oberman et al., 2005, 2008; but see Raymaekers et al., 2009). Most relevant to social interaction skills is an experiment by Oberman et al. (2007) in which the authors show modulation of mu suppression as a function of the presumed involvement of the participant in a virtual ball game carried on in a video clip (i.e., watching a video of players play separately with a ball, play with each other, or throw the ball towards the participant). However, it should be noted that mu rhythms oscillate in the same frequency range as the well known alpha rhythms which culminate over the occipital cortex (Berger, 1929; for a review see Klimesch, 1999). The alpha frequency dominates the EEG when the brain rests ("idling rhythms"; e.g., Pfurtscheller et al., 1996). Suppression of alpha waves is thought to reflect enhancement of neural activity induced by a perceptual event, which leads to asynchronous neural firing. An overall decrease in alpha power has been also linked, however, to increasing demands of attention, alertness, episodic memory, and task-load in general (for reviews see Klimesch, 1999; Sauseng & Klimesch, 2008). Since perceptual studies involving biological motion of a target may capture participants' attention more than nonbiological stimuli, alpha and mu suppression are sometimes modulated in parallel and thus these two manifestations are hard to disentangle. Therefore, when studying mu suppression, it is important to examine and report effects in both occipital and central (sensorymotor) regions (cf., Perry & Bentin, 2010). In the present experiment we extended Oberman et al.'s idea of an interactive game, asking participants to watch, imagine playing and actually play a Rock–Paper–Scissors (RPS) game. We compared these conditions to conditions in which no social interaction was necessary, that is, participants were asked to act and view actions out of the context of this game. The Rock–Scissors–Paper game is particularly interesting, since it involves not only a social interaction, but also competition. Humans are known to be extremely sensitive to

competition and social comparison (Dvash et al., 2010), thus enhancing the participants' involvement in the task. Consistent with previous studies, we expected to find both alpha and mu suppression in the viewing of biological motion segments of each condition (e.g. Perry & Bentin, 2010; Perry et al., 2010b), but primarily mu suppression in the action conditions. In addition, we expected to find more mu suppression, indexing more motor system involvement, as the social interaction increases, that is, mostly when the participants view a game in which they actually play and the least when the movements are seen out of context. The rationale behind this hypothesis is that when one is actually playing a social game, the opponent's moves directly affect the participant's winning of the game. Thus, the participant should be most engaged in deciphering these moves and, in those circumstances, we expect the highest recruitment of a motor mirror-like system. In situations in which the participant is only a viewer of the game or sees random gestures, an automatic mirror system might be recruited, but to a lesser degree. We also expected that this trend will be different from the trend seen in the alpha (occipital) rhythms, where more suppression should be correlated with more visual stimuli, regardless of motor or social context. Materials and methods Participants Twenty eight participants (14 male) took part in the study. All were undergraduate students from the Hebrew University ranging in age from 20 to 30 years (mean age 24.6, SD = 2.4). They participated in the experiment for course credit or payment. One participant was left-handed. All participants reported normal or corrected to normal visual acuity and had no history of psychiatric or neurological disorders (confirmed by a screening interview). All participants grew up in Israel insuring that they were familiar with the RPS game, which is a very well known and common game in Israel. Stimuli task and design The stimuli used were 2 s long color video clips presenting a right hand performing the three acts used in the game—"rock", "scissors" or "paper", from either the direction of the participant or facing the participant. The length of the movement itself was 1 second, starting about 200 ms from the onset of the clip (Fig. 1). For the "view no context" block (see below) we also used a video clip of the same right hand doing a "thumbs-up" sign. The “rolling balls” baseline (see below) was composed of video clips presenting a tennis ball rolling from either the right or the left side of a table. The video clips were presented on a CRT monitor, 70 cm away from the participant's eyes with the hand displays subtending on average a visual angle of 3 × 8°. Presentation (Neurobehavioral Systems) was used for stimulus presentation and experimental control. In a blocked design, each block depicted a different task. The first block, "view–no context", depicted 100 clips of "rock", "scissors" or "paper" gestures presented in random order from the participant's angle and facing the participant in alternation (as if in a RPS game). Among these clips, we introduced 8–12 clips of the "thumbs-up" sign, and participants were instructed to silently count the "thumbs-up" clips and report their number of the end of the block. In the second block, "view–game", the same 100 clips were shown. Participants were told that they will be watching a RPS game, and were reminded of the rules of the game ("rock" beats "scissors", "scissors" beat "paper" and "paper" beats "rock"). Participants were asked to count how many times one of the two players won (counterbalanced between participants). The experimental control program was set to make the target player win between 8 and 12 times. In the third block, "view–imagine", participants were told that they were now a player in

A. Perry et al. / NeuroImage 58 (2011) 895–904

897

Fig. 1. Examples of the stimuli presented in the experiment - across from the participant (top row) and from the participant's angle (bottom row). From left to right: rock, scissors, and paper.

the game, and each time a cross appeared on the screen, they should imagine performing one of the three game movements with their right hand. Two seconds following the cross, the opponent's move was shown on the screen. Altogether there were 50 such "games". In all the above conditions, participants were instructed to refrain from any actual hand movement and were monitored by camera throughout the experiment to make sure they complied with the instructions. In the fourth block, "view–play", the same instructions were given as in the previous block, except that participants were instructed to actually perform the game acts, with a small right-palm movement, on a small board placed on their knees (this was done to avoid whole arm movements which may disturb the EEG recordings). Each time a cross appeared on the screen, participants made their move. Two seconds later the opponent's move was shown on the screen, facing the participant. Participants were told that although this is not the natural way to play, their opponent does not see their move in advance. The fifth block, "balls–act", served to isolate the game from the movement component in the fourth block; in this block participants saw a cross and actually performed one of the RPS moves as before; however, after 2 s, a ball appeared on the screen, rolling either from right to left or from left to right. In this block participants were instructed to count how many times the ball rolled from one of the directions to the other (counterbalanced between participants). The cross and the balls appeared 50 times each, and the ball rolling in the target direction appeared between 8 and 12 times. The sixth block, served as a baseline for the rest of the blocks, depicting non-biological motion only. Participants viewed a ball rolling either from right to left or from left to right, and were instructed to count how many times the ball rolled from one of the directions to the other (counterbalanced between participants). One direction appeared 100 times, and the direction which the participant was instructed to count appeared between 8 and 12 times. The blocks were presented in the order described above, except for the baseline block (block six) which appeared either first or last, counterbalanced between participants. Data acquisition and analysis EEG recording The EEG analog signals were recorded continuously (from DC with a low-pass filter set at 100 Hz) by 64 Ag–AgCl pin-type active

electrodes mounted on an elastic cap (Biosemi TM, http://www. biosemi.com/headcap.htm) according to the extended 10–20 system, and from two additional electrodes placed at the right and left mastoids. All electrodes were referenced during recording to a common-mode signal (CMS) electrode between POz and PO3 and were subsequently re-referenced digitally (see data processing below). Eye movements, as well as blinks, were monitored using bipolar horizontal and vertical EOG derivations via two pairs of electrodes, one pair attached to the external canthi, and the other to the infraorbital and supraorbital regions of the right eye. Both EEG an EOG were digitally amplified and sampled at 256 Hz using a Biosemi Active II system (www.biosemi.com). Data processing Data were analyzed using Brain Vision Analyzer software (Brain Products; www.brainproducts.com) and Matlab routines. Raw EEG data was initially high-pass filtered at 0.5 Hz (24 dB) and rereferenced off-line to the digital average of the two mastoids. EEG deflections resulting from eye movements and blinks were corrected using an ICA procedure (Jung et al., 2000). Remaining artifacts exceeding ±100 μV in amplitude were rejected. We divided the conditions into perception segments and action segments. In blocks that depicted two different alternating "events" (e.g. playing and then viewing a player) we distinguished between perception and action conditions. In blocks that depicted only perceived events, we took only the fifty trials which were similar to those viewed in the other conditions. This provided us with a total of four perception conditions, three action conditions and a baseline condition. The perception conditions were: view in the no context condition (VNoContext), view in the game condition (VGame), view in the imagine condition (VImagine), and view in the play condition (VPlay). The action conditions consisted of Imagine and Play from the relevant conditions, and Act taken from the last block where participants performed the movements, without the context of a game. During all three events, a cross appeared at the center of the screen. From each 2-second video segment, we analyzed only the 1-second segment in which there was actual motor movement. For the perception conditions, this was from 200 ms after the clip's onset, and for segments in which the participants themselves performed an act, we took a second-long segment from the appearance of the cross signaling to start the

898

A. Perry et al. / NeuroImage 58 (2011) 895–904

motion, as this was the average time the movement took (see also wavelet analysis and distributions bellow). For each of the 1-second segments, the integrated power in the 8–13 Hz range was computed using a Fast Fourier Transform (FFT) performed at 0.5 Hz intervals (using a Hanning window). The segments were then averaged for each condition. A suppression index was calculated as the logarithm of the ratio of the power during each condition relative to the power during the baseline condition, and used as dependent variable. The ratio (as opposed to a simple subtraction) was used to control for the variability in absolute EEG power as a result of individual differences such as scalp thickness and electrode impedance (Pineda & Oberman, 2006). The log transform was applied to the ratio before statistical analyses because ratio data are inherently not normally distributed as a result of lower bounding. A log ratio of less than zero indicates suppression in the EEG amplitude, whereas a value of zero indicates no change and values greater than zero indicate enhancement. Suppression was computed at 2 occipital sites, O1 and O2, where alpha modulation is expected; and 2 central sites C3 and C4, which are assumed to be mu rhythm sites. In order to validate the choice of the 8–13 Hz band, and the timing of the actions, we also performed a wavelet analysis. We performed the wavelet analysis on single trials at each recording site in order to obtain the power of the EEG activity at all frequencies ranging from 1 to 20 Hz at different time points. We used a complex Gaussian Morlet wavelet with width of the wavelet determined according to the Morlet parameter 5, and in steps of 1 Hz. The motivation for choosing a relatively low Morlet parameter was to improve the temporal resolution of these frequencies which tend to be “smeared” over time (Zion-Golumbic et al., 2010).We averaged the amplitudes at each time–frequency point at each recording site across trials for each subject in each condition. We then calculated the suppression index for each point, similarly to the index calculated for the FFT analysis, as the logarithm of the ratio of the power during each condition relative to the power during the baseline condition, for each point. Suppression was computed at the same sites as the FFT, that is: occipital sites O1 and O2, and central sites C3 and C4. Results Although gender differences in the expression of the hMNS have been occasionally reported in fMRI studies (e.g., Schulte-Rüther et al., 2008), we did not find such differences in the present study. Therefore, the data were collapsed and analyzed across gender. Perception conditions The suppression index was analyzed using ANOVA with repeated measures (Bonferroni corrected wherever multiple comparisons were made). The factors were Region (Central, Occipital), Hemisphere (Left, Right), and Condition (VNoContext, VGame, VImagine, VPlay). The degrees of freedom were corrected using the Greenhouse–Geisser epsilon values (G–GE) when needed. Relative to the “rolling balls” baseline, the amplitude of the 8– 13 Hz rhythms was significantly suppressed in all conditions, at occipital as well as at central sites. However, the pattern of suppression was different at the posterior and more anterior sites (Fig. 2). ANOVA showed significant main effects for Region [F(1,27) = 14.9, MSe = 0.238, p b 0.001] and Condition [F(3, 81) = 3.0, MSe = 0.168, p b 0.05]. However, these main effects were qualified by a second-order interaction of Hemisphere × Condition [F(3, 81) = 7.1, MSe = 0.01, p = 0.001] and by a third order interaction of Region × Hemisphere × Condition [F(3, 81) = 7.4, MSe = 0.007, p b 0.001]. This interaction was further investigated by separate Hemisphere × Condition ANOVAs for each region.

In the occipital region there was a main effect of Condition [F(3,81)] = 3.1, MSe = 0.124, p b 0.05], qualified by a significant interaction between Hemisphere × Condition [F(3,81) = 3.4, MSe = 0.008, p b 0.005]. Pairwise comparisons, showed that in the right occipital site (O2) there was a significant difference between VGame [Mean (M) = −0.49] and VNoContext (M = −0.38; p b 0.01), and between VGame and VImagine (M = −0.28, p b 0.01), and no significant differences between these and VPlay (M = −0.40). In the left occipital site (O1) there was only a significant difference between VGame (M = −0.46) and VNoContext (M = −0.36; p b 0.05) and not between these and the other two conditions (M VPlay = −0.43; M VImagine = −0.33; Fig. 2a). In the central region there were significant main effects for Hemisphere [M Left = − 0.25, M Right = − 0.19; F(1,27) = 7.3, MSe = 0.031, p b 0.05], and a significant interaction between Condition × Hemisphere [F(3,81) = 10.6, MSe = 0.01, P b 0.001]. Pairwise comparisons, showed that in the left central site (C3) there was a significant difference between VPlay (M = −0.38) and VImagine (M = −0.16, p b 0.01), VPlay and VNoContext (M = − 0.23, p b 0.05) and a close to significant difference between VPlay and VGame (M = −0.23, p = 0.07). However, after Bonferroni correction, only the first difference (VPlay–VImagine) was significant. There were no significant differences between conditions in the right central site (C4; M VPlay = − 0.20; M VImagine = − 0.15;M VGame = − 0.22; M VNoContext = − 0.18; Fig. 2b). The topographical distributions depict the log of the ratio between each condition and baseline. The distributions show that in the viewing right-hands segments, suppression in the 8–13 Hz was evident all over the scalp, but was strongest in occipital regions bilaterally, thus probably conveying mostly alpha suppression. The VPlay condition differs in this respect, as suppression is also strong in the central region and tilted over the left hemisphere (Fig. 2c). Action conditions As in the View conditions, the suppression index was analyzed using ANOVA with repeated measures. The factors were Region (Central, Occipital), Hemisphere (Left, Right), and Condition (Imagine, Play, Act). Recall that in the Play and Act segments participants were asked to do the exact same RPS movements, however, whereas in the Play condition they actually interacted with and “opponent”, in the Act condition their move was not reciprocated followed only by rolling balls. As evident in Fig. 3a, mu suppression emerged in the Play and ACT conditions, but no suppression occurred during the imagine condition. ANOVA showed a significant effect of Region [M Central = − 0.15, M Occipital = 0.12; F(1,27) = 36.3, MSe = 0.163, p b 0.001], Hemisphere [M Left = −0.50, M Right = 0.01; F(1,27) = 7.5, MSe = 0.05, p b 0.05] and Condition [F(2, 54) = 9.5, MSe = 0.16, p b 0.001], qualified by a Hemisphere × Condition interaction [F(2,54) = 4.4, MSe = 0.064, p b 0.05]. Although in both occipital and central regions there was a similar condition effect (and thus there was no Region × Condition interaction), in the occipital region the action conditions show an increase in power relative to baseline (i.e. synchronization), while in the central region there is clear suppression (i.e., desynchronization). Since our interest and predictions were focused on suppression effects, we went on to analyze the Hemisphere × Condition interaction in the central region only, using pairwise comparisons (Bonferroni corrected). In the left hemisphere (C3) there were significant differences between Imagine and Act (M = 0.007 and − 0.17 respectively; p = 0.01) and between Imagine and Play (M = −0.36; p b 0.001), and a difference reaching significance between Play and Act (p = 0.058). In the right hemisphere (C4) there were significant differences between Imagine and Play (M = −0.014 and − 0.24; p = 0.001) and between Play and Act (M = −0.12, p b 0.05), but not between Imagine and Act.

A. Perry et al. / NeuroImage 58 (2011) 895–904

(a)

(c)

Occipital Electrodes

FFT suppression index

O1

O2

VNo Context

VGame

VImagine

VPlay

0 -0.1 -0.2 -0.3 -0.4 -0.5 -0.6

*

*

VNoContext

VGame

(b)

*

VImagine

VPlay

Central Electrodes C3 FFT suppression index

899

C4

0 -0.1 -0.2 -0.3 -0.4 -0.5

*

-0.6 VNoContext

VGame

VImagine

VPlay -0.5

0

0.1

Fig. 2. Suppression for the perception conditions in (a) the occipital region (b) the central region. The Y axis indicates the FFT suppression index (log ratio of the power in the experimental conditions over baseline; a value of zero means no suppression, see text). (c) The distribution of the different perception conditions, all depicting the same RPS stimuli.

The topographical distribution of the Play condition compared to baseline shows clear suppression in the central region, with enhancement of power (synchronization) in the occipital region.

(a) FFT suppression index

0.1

C3

C4

p = 0.058

*

0 -0.1 -0.2 -0.3

*

-0.4 -0.5

*

* Imagine

(b)

The Act condition conveys a similar central distribution to that of Play. The Imagine condition does not show suppression compared to baseline (Fig. 3b).

Imagine

Act

Play

Act

-0.5

Play

0

0.1

Fig. 3. (a) Suppression for the three action conditions in the central region. The Y axis indicates the FFT suppression index (log ratio of the power in the experimental conditions over baseline; a value of zero means no suppression, see text). (b) The distribution of the different action conditions.

900

A. Perry et al. / NeuroImage 58 (2011) 895–904

Wavelet analysis The wavelet analysis clearly showed that the suppression was most dominant at the frequency range of 8–13 Hz as we expected. In addition, this analysis revealed that the main effect of the suppression during the perception conditions was mainly during 0.2–1.2 s from the stimulus onset, matching the time the hand movement appeared on the screen (see Fig. 4). During the action conditions, the suppression occurred at the first second of the condition, matching the time the subjects performed the required action (see Fig. 5). A noticeable difference between hemispheres is shown at the central sites during both the Play (Fig. 5c) and the VPlay (Fig. 4d) conditions. It is apparent that the suppression at the left hemisphere site (contralateral to the moving hand) was greater than the suppression at the right hemisphere site. A similar hemisphere effect is seen for VImagine in the occipital region. The hemispheric difference is also evident in the FFT distributions and graphs. In order to investigate this effect further, for each region, we calculated a hemispheric difference index for each condition (C3–C4 and O1–O2), and analyzed the differences between the conditions using Post-hoc Bonferroni contrasts. In the central region, the hemispheric difference index of VPlay (M Central Difference index = −0.17) was indeed significantly different from all other perception conditions (M Central Difference indexes: VImagine = − 0.01, VGame = − 0.02, VNoContext = − 0.05; p b 0.05 for all), with no other significant differences between them. In the action conditions, no statistically significant differences between conditions was found (M Central Difference indexes: Play = −0.12, Act = −0.05, Imagine = 0.02). In the Occipital region, there was only a significant difference between the VImagine and VGame condition (pb 0.05), with no other significant differences between any of the conditions (M Occipital Difference indexes: VPlay = −0.03; VImagine= −0.05, VGame = 0.03, VNoContext= 0.02; Play = −0.09, Act = −0.09, Imagine= −0.06). A further important outcome of the wavelet analysis is the evident enhancement of mu and alpha waves after the end of the movement (execution or imagination) during the action conditions. This enhancement, also known as a "rebound effect" is in accordance with previous MEG experiments (see discussion). Discussion In the present study we investigated the effects of different levels of social interaction on mu and alpha suppression (8–13 Hz), which are presumed to index motor-mirror and attentional mechanisms respectively. In the central left site, we found greater mu suppression in both the perception and the acting segments of the Play condition than in other conditions in which the participants were not actually involved in a game interaction. That is, when participants actually played the RPS game against an opponent, the amplitude of the EEG oscillations in the mu range was reduced both while acting and while watching the opponent act, suggesting somatosensory and motor system involvement in both conditions. This effect was unique to the central mu region and it was lateralized to the left hemisphere, contralateral to the observed playing hand as well as to the participant's own moving hand. This result supports the hypothesis that motor (and maybe mirror) mechanisms are affected by the level of the participant's personal involvement and by the social context of the perceived motion. It has been recently suggested that the recruitment of the hMNS while viewing a competitive game may reflect the participant's internal assimilation of the actions of the observed figures (Shimada & Abe, 2010). The present data suggest that such assimilation becomes more important as one becomes more involved in the task, that is, while actually playing rather than only viewing the game. It should be noted, however, that in the perception segments the effect was small, and significant only when compared to the viewing segments in the Imagine condition.

Conversely, the occipital region exhibited the greatest suppression for the VGame condition, the first condition in which the game is explicitly introduced to the participants. In the central (mu) region the level of EEG suppression in the VGame condition was similar to that in the VNoContext condition, in contrast to our initial expectation. This dissociation shows that both alpha and mu suppression are affected by social aspects, but in different ways. While mu suppression was enhanced in the perception segments only when the participant was actually part of the interactive game, the occipital suppression might reflect the recruitment of attention resources, engaged by introducing the game itself (Klimesch, 1999). A major outcome of the present study is the significantly larger mu suppression while the participants actually played the game against the virtual opponent on the screen (Play condition) than while they performed the same acts except that there was no opponent to play back (Act condition). The greater suppression in the Play condition may signal greater involvement of motor regions as a factor of the participants' engagement in social interaction. It should be noted that although seeing videos of real hand actions, participants usually knew that they were playing "against the computer", rather than against a "rational agent" (see Gallagher et al., 2002). Playing against a real human player might, indeed, enhance this effect even more, adding a more explicit component of mentalizing in order to infer the other's beliefs and intentions which may hint to the opponent's future moves (Gallagher et al., 2002; see also Lee et al., 2005 for a study showing such decision making strategies in monkeys while playing the RPS game). Additional mu-experiments in which two human participants play one against the others are necessary to examine this assumption. The action conditions also showed a sharp contrast between mu and alpha. Indeed, in the occipital region there was no suppression of EEG in the 8–13 Hz when averaged in FFT, and a much smaller and shorter effect when viewed in the wavelet analysis. Another finding that distinguishes the central region from the occipital region is the different spatial suppression pattern. In both the Play and the VPlay conditions we see a tendency towards a left lateralized pattern in the central regions and not in the occipital regions. We believe that this lateralization of mu suppression is due to greater motor-attention focused on the right hand, both of the participant playing, and of the opponent, who also played with her right hand (see also Perry and Bentin 2009 for lateralization of mu suppression following contralateral hand movements). In the occipital region, there is a left lateralized tendency in the VImagine condition only. Because of its distribution, this suppression probably does not depict lateralized motor simulation, but rather the recruitment of general attentional resources. These distribution differences may indicate different inner processes underlying the suppression in the central and occipital regions. The post-movement "rebounds" (or enhancements) seen in the wavelet analysis, are similar to those previously described in MEG, following hand movements or median-nerve stimulation (Salmelin and Hari, 1994; Schnitzler et al., 1997). However, while Schnitzler et al. suggest that this effect reflects changes in the functional state of the sensory-motor cortex, we see these rebounds in both occipital and central regions. There was no EEG suppression in the central region while the participants imagined playing the game (that is imagined an act). The lack of central mu suppression in the Imagine condition is surprising in light of the repeated demonstration of mu suppression over the sensory-motor cortex in imagination experiments (e.g. Pfurtscheller & Neuper, 1997; Pfurtscheller et al., 2006, 2008; Pineda et al., 2000; Schnitzler et al., 1997). Perhaps the absence of mu suppression in our task reflects a jitter in the onset of imaging which has been prevented in the studies of Pruftscheller's group. Whereas in their studies the imagined action was initiated by the experimenter by presenting its name, in our study participants first made a decision and then imagined the execution. Hence it is possible that in different trials the

(a)

(b)

VNoContext

10

10

5

5

0

0

1

1.5

2

0.4

0.3

0.2 0.5

t [sec] O1

1.5

2

O2 20

15

15

f [Hz]

20

10 5

15

10

5

0

0

1

1.5

2

0.3

0.2 0.5

t [sec]

5

1

1.5

2

t [sec]

O1

-0.1 -0.2

0.5

0.4

10

5

0.1

0

10

15

O2

20

20

15

15

10

0

-0.1

10

5

-0.2

5 -0.3

-0.3 1.5

0

2

0.5

t [sec]

20

15

15

10

5

0

0

1.5

2

0.4

0.3

0.2 0.5

1

1.5

2

t [sec]

O1

O2 20

15

15

f [Hz]

20

10 5

1.5

0

2

0.5

t [sec]

10

5

1

15

15

10

5

0

0

-0.2

1.5

2

0.4

0.3

0.2 0.5

1

1.5

2

t [sec]

O1

-0.1

5

1

2

10

5

0.5

1.5

C4

20

t [sec]

0

10

VPlay

C3

20

0.1

1

t [sec]

f [Hz]

20

f [Hz]

f [Hz]

0.5

C4

t [sec]

f [Hz]

0

2

(d)

VImagine

C3

1

1.5

t [sec]

(c)

0.5

1

O2

20

20

15

15

f [Hz]

1

f [Hz]

0.5

f [Hz]

0

10 5

0.5

1

t [sec]

1.5

2

0

0.5

1

t [sec]

1.5

2

0.1

0

-0.1

10

-0.2

5 -0.3

-0.3 0

0.1

A. Perry et al. / NeuroImage 58 (2011) 895–904

f [Hz]

1

t [sec]

20

f [Hz]

15

C4

20

f [Hz]

15

f [Hz]

20

f [Hz]

20

0.5

VGame C3

C4

f [Hz]

f [Hz]

C3

0

0.5

1

t [sec]

1.5

2

0

0.5

1

1.5

2

t [sec]

Fig. 4. Suppression of the Morlet wavelet power for the perception conditions in the four sites (O1, O2, C3, C4). (a) VNoContext, (b) VGame, (c) VImagine, (d) VPlay. The X axis indicates time (seconds); The Y axis indicates frequency (Hz); The colors indicate the suppression index (log ratio of the power in the experimental conditions over baseline).

901

902

A. Perry et al. / NeuroImage 58 (2011) 895–904

(a)

Imagine C4

C3

20

20 0.4

15

f [Hz]

f [Hz]

15 10 5

5

0

0

0.5

1

1.5

2

0.3

10

0.2 0.5

t [sec] O1

1.5

2 0.1

O2

20

20

15

15

f [Hz]

f [Hz]

1

t [sec]

10 5

0

-0.1

10

-0.2

5 -0.3

0

0.5

1

1.5

0

2

0.5

t [sec]

C4 20

15

15

10

10

5

5

0

0

0.5

1

1.5

2

0.4

0.3

0.2 0.5

t [sec]

1.5

2 0.1

O2

20

20

15

15

f [Hz]

f [Hz]

1

t [sec]

O1

10 5

0

-0.1

10 -0.2 5

0.5

1

1.5

0

2

-0.3 0.5

t [sec]

1

1.5

2

t [sec]

(c)

Play C3

20

C4

20

0.4

15

f [Hz]

15

f [Hz]

2

Act C3

20

0

1.5

t [sec]

f [Hz]

f [Hz]

(b)

1

10 5

10

0.3

5 0.2

0

0.5

1

1.5

0

2

0.5

t [sec] O1

2 0.1

O2 20

15

15

f [Hz]

f [Hz]

1.5

t [sec]

20

10 5 0

1

0

-0.1

10 -0.2 5

0.5

1

t [sec]

1.5

2

0

-0.3 0.5

1

1.5

2

t [sec]

Fig. 5. Suppression of the Morlet wavelet power for the action conditions in the four sites (O1, O2, C3, C4). (a) Imagine, (b) Act, (c) Play. The X axis indicates time (seconds); The Y axis indicates frequency (Hz); The colors indicate the suppression index (log ratio of the power in the experimental conditions over baseline).

A. Perry et al. / NeuroImage 58 (2011) 895–904

onset of imagination varied and, averaging across trials the effect was wiped away. This account is particularly feasible since the action imagery effect described by Pfurtscheller and Neuper (1997) was a short lasting (~500 ms), not evident in all participants and varying in the suppressed frequency range (Pfurtscheller et al., 2008). In addition, it is also noteworthy that, although instructed to imagine the move itself, many participants reported finding this task difficult to do, and instead imagined the word or concept, and not the hand movement. It should be noted that the low spatial resolution of EEG (and in this respect of all neuroimaging techniques) makes it impossible to infer that the same neurons are firing in the action and viewing conditions; all we can say is that the same region / system is involved (Fabbri-Destro & Rizzolatti, 2008). To this end, it is worth mentioning a study by Dinstein et al. (2008). Also using the RPS game and recording brain activations by fMRI, these authors showed that the spatial pattern of activity for a particular movement in the anterior intraparietal sulcus (aIPS) was distinctly different when it was only observed than, when the same movement was executed, by the participant. The different patterns of activity in the same aIPS region suggest that observed and executed movements are probably represented by distinctly different subpopulations of neurons in aIPS. Similarly, although mu suppression has been frequently considered an index of mirror neuron activity (Muthukumaraswamy, et al., 2004; Oberman, et al., 2005, 2007), the resolution of EEG does not enable to differentiate between activity selective to the sensorymotor regions and activity in other regions which, together, may form a larger action observation/ execution network, modulating the activity in sensory-motor areas. Future research with higher-spatialresolution techniques and statistical algorithms, should continue the efforts already made to further dissociate between different sources of activation and confirm at what stage of processing and where in the brain the information regarding social content is processed (Hari, et al., 1997; Hari et al., 2000; Hyvarinen et al., 2010). To conclude, this study explored the effect of different levels of social interaction on EEG mu and alpha rhythms, indexing a possible human analogue of the mirror-motor system and a perceptual– attentional mechanism, respectively. We found tentative support to the hypothesis that mu suppression is affected by the social context, that is, that more engagement in a social game elicits greater and lateralized mu suppression. This interpretation is in line with previous studies that showed greater mu suppression in tasks involving social interactions between self and others (Oberman, et al., 2007) and understanding others' intended direction of movement (towards or away from oneself; Perry, et al., 2010b). However, we did not find differences in mu suppression between perception segments in which the participant did not actually play, and occipital alpha suppression was more robust in all perception conditions. Acknowledgments We thank Shirey Sole and Noga Diamant for skillful assistance in running the experiment. Anat Perry was partially funded by the Noah Royal Foundation and the “Hoffman Leadership and Responsibility” fellowship program, at the Hebrew University. References Adolphs, R., 2003. Cognitive neuroscience of human social behaviour. Nat. Rev. Neurosci. 4 (3), 165–178. Agnew, Z.K., Bhakoo, K.K., Puri, B.K., 2007. The human mirror system: a motor resonance theory of mind-reading. Brain Res. Rev. 54 (2), 286–293. Berger, H., 1929. Uber das Elektrenkephalogramm des Menschen. Arch. Psychiatr. Nervenkr. 87, 527–570. Buccino, G., Binkofski, F., Fink, G.R., Fadiga, L., Fogassi, L., Gallese, V., et al., 2001. Action observation activates premotor and parietal areas in a somatotopic manner: an fMRI study. Eur. J. Neurosci. 13 (2), 400–404.

903

Carr, L., Iacoboni, M., Dubeau, M.C., Mazziotta, J.C., Lenzi, G.L., 2003. Neural mechanisms of empathy in humans: a relay from neural systems for imitation to limbic areas. Proc. Natl. Acad. Sci. U. S. A 100 (9), 5497–5502. Carrington, S.J., Bailey, A.J., 2009. Are there theory of mind regions in the brain? A review of the neuroimaging literature. Hum. Brain Mapp. 30 (8), 2313–2335. Carruthers, P.a.S.,.P., 1996. Theories of Theories of Mind. Cambridge University Press. Cheng, Y., Yang, C.-Y., Lin, C.-P., Lee, P.-L., Decety, J., 2008. The perception of pain in others suppresses somatosensory oscillations: a magnetoencephalography study. NeuroImage 40 (4), 1833–1840. Davis, M., Stone, T. (Eds.), 1995. Mental Simulation. Blackwell, Oxford, UK. Decety, J., Chaminade, T., Grezes, J., Meltzoff, A.N., 2002. A PET exploration of the neural mechanisms involved in reciprocal imitation. NeuroImage 15 (1), 265–272. di Pellegrino, G., Fadiga, L., Fogassi, L., Gallese, V., Rizzolatti, G., 1992. Understanding motor events: a neurophysiological study. Exp. Brain Res. 91 (1), 176–180. Dinstein, I., Gardner, J.L., Jazayeri, M., Heeger, D.J., 2008. Executed and observed movements have different distributed representations in human aIPS. J. Neurosci. 28 (44), 11231–11239. Dvash, J., Gilam, G., Ben-Ze'ev, A., Hendler, T., Shamay-Tsoory, S.G., 2010. Envious brain: the neural basis of social comparison. Hum. Brain Mapp. 31, 1741–1750. Fabbri-Destro, M., Rizzolatti, G., 2008. Mirror neurons and mirror systems in monkeys and humans. Physiology 23 (3), 171–179. Fadiga, L., Fogassi, L., Pavesi, G., Rizzolatti, G., 1995. Motor facilitation during action observation: a magnetic stimulation study. J. Neurophysiol. 73 (6), 2608–2611. Gallagher, H.L., Jack, A.I., Roepstorff, A., Frith, C.D., 2002. Imaging the intentional stance. NeuroImage 16, 814–821. Gallese, V., Fadiga, L., Fogassi, L., Rizzolatti, G., 1996. Action recognition in the premotor cortex. Brain 119 (Pt 2), 593–609. Gastaut, H., 1952. Electrocorticographic study of the reactivity of rolandic rhythm. Rev. Neurol. 87 (2), 176–182 (Paris). Grezes, J., Armony, J.L., Rowe, J., Passingham, R.E., 2003. Activations related to "mirror" and "canonical" neurons in the human brain: an fMRI study. NeuroImage 18 (4), 928–937. Happe, F., Ehlers, S., Fletcher, P., Frith, U., Johansson, M., Gillberg, C., et al., 1996. 'Theory of mind' in the brain. Evidence from a PET scan study of Asperger syndrome. Neuroreport 8 (1), 197–201. Hari, R., Kujala, M.V., 2009. Brain basis of human social interaction: from concepts to brain imaging. Physiol. Rev. 89 (2), 453–479. Hari, R., Levanen, S., Raij, T., 2000. Timing of human cortical functions during cognition: role of MEG. Trends Cogn. Sci. 4 (12), 455–462. Hari, R., Salmelin, R., Makela, J.P., Salenius, S., Helle, M., 1997. Magnetoencephalographic cortical rhythms. Int. J. Psychophysiol. 26 (1–3), 51–62. Hyvarinen, A., Ramkumar, P., Parkkonen, L., Hari, R., 2010. Independent component analysis of short-time Fourier transforms for spontaneous EEG/MEG analysis. NeuroImage 49 (1), 257–271. Iacoboni, M., 2009. Imitation, empathy, and mirror neurons. Annu. Rev. Psychol. 60, 653–670. Iacoboni, M., Dapretto, M., 2006. The mirror neuron system and the consequences of its dysfunction. Nat. Rev. Neurosci. 7 (12), 942–951. Jackson, P.L., Decety, J., 2004. Motor cognition: a new paradigm to study self-other interactions. Curr. Opin. Neurobiol. 14 (2), 259–263. Jung, T.P., Makeig, S., Westerfield, M., Townsend, J., Courchesne, E., Sejnowski, T.J., 2000. Removal of eye activity artifacts from visual event-related potentials in normal and clinical subjects. Clin. Neurophysiol. 111 (10), 1745–1758. Keysers, C., Perrett, D.I., 2004. Demystifying social cognition: a Hebbian perspective. Trends Cogn. Sci. 8 (11), 501–507. Klimesch, W., 1999. EEG alpha and theta oscillations reflect cognitive and memory performance: a review and analysis. Brain Res. Rev. 29 (2–3), 169–195. Lee, D., McGreevy, B.P., Barraclough, D.J., 2005. Learning and decision making in monkeys during a rock–paper–scissors game. Cogn. Brain Res. 25 (2), 416–430. Leslie, K.R., Johnson-Frey, S.H., Grafton, S.T., 2004. Functional imaging of face and hand imitation: towards a motor theory of empathy. NeuroImage 21 (2), 601–607. Lieberman, M.D., 2007. Social cognitive neuroscience: a review of core processes. Annu. Rev. Psychol. 58, 259–289. Martineau, J., Schmitz, C., Assaiante, C., Blanc, R., Barthelemy, C., 2004. Impairment of a cortical event-related desynchronisation during a bimanual load-lifting task in children with autistic disorder. Neurosci. Lett. 367 (3), 298–303. Mukamel, R., Ekstrom, A.D., Kaplan, J., Iacoboni, M., Fried, I., 2010. Single-neuron responses in humans during execution and observation of actions. Curr. Biol. 20 (8), 750–756. Muthukumaraswamy, S.D., Johnson, B.W., McNair, N.A., 2004. Mu rhythm modulation during observation of an object-directed grasp. Brain Res. Cogn. Brain Res. 19 (2), 195–201. Niedenthal, P.M., Barsalou, L.W., Winkielman, P., Krauth-Gruber, S., Ric, F., 2005. Embodiment in attitudes, social perception, and emotion. Pers. Soc. Psychol. Rev. 9 (3), 184–211. Oberman, L.M., Hubbard, E.M., McCleery, J.P., Altschuler, E.L., Ramachandran, V.S., Pineda, J.A., 2005. EEG evidence for mirror neuron dysfunction in autism spectrum disorders. Brain Res. Cogn. Brain Res. 24 (2), 190–198. Oberman, L.M., Pineda, J.A., Ramachandran, V.S., 2007. The human mirror neuron system: a link between action observation and social skills. Soc. Cogn. Affect. Neurosci. 2 (1), 62–66. Oberman, L.M., Ramachandran, V.S., Pineda, J.A., 2008. Modulation of mu suppression in children with autism spectrum disorders in response to familiar or unfamiliar stimuli: the mirror neuron hypothesis. Neuropsychologia 46 (5), 1558–1565. Perry, A., Bentin, S., 2009. Mirror activity in the human brain while observing hand movements: equivalence between EEG desynchronization in the μ-range and fMRI. Brain Res. 1282, 126–132.

904

A. Perry et al. / NeuroImage 58 (2011) 895–904

Perry, A., Bentin, S., 2010. Does focusing on hand-grasping intentions modulate electroencephalogram [mu] and [alpha] suppressions? Neuroreport. Perry, A., Bentin, S., Ben-Ami Bartal, I., Lamm, C., Decety, J., 2010a. “Feeling” the pain of those who are different from us—modulation of EEG in the mu/alpha range. Cogn. Affect. Behav. Neurosci. 10, 493–504. Perry, A., Troje, N.F., Bentin, S., 2010b. Exploring motor system contributions to the perception of social information: evidence from EEG activity in the mu/alpha frequency range. Soc. Neurosci. 5 (3), 272–284. Pfeifer, J.H., Lieberman, M.D., Dapretto, M., 2007. "I know you are but what am I?!": neural bases of self- and social knowledge retrieval in children and adults. J. Cogn. Neurosci. 19 (8), 1323–1337. Pfurtscheller, G., Brunner, C., Schlogl, A., Lopes da Silva, F.H., 2006. Mu rhythm (de) synchronization and EEG single-trial classification of different motor imagery tasks. NeuroImage 31 (1), 153–159. Pfurtscheller, G., Neuper, C., 1997. Motor imagery activates primary sensorimotor area in humans. Neurosci. Lett. 239 (2–3), 65–68. Pfurtscheller, G., Scherer, R., Muller-Putz, G.R., da Silva, F.H.L., 2008. Short-lived brain state after cued motor imagery in naive subjects. Eur. J. Neurosci. 28 (7), 1419–1426. Pfurtscheller, G., Stancak, A., Neuper, C., 1996. Event-related synchronization (ERS) in the alpha band—an electrophysiological correlate of cortical idling: a review. Int. J. Psychophysiol. 24 (1–2), 39–46. Pineda, J.A., 2005. The functional significance of mu rhythms: translating "seeing" and "hearing" into "doing". Brain Res. Rev. 50 (1), 57–68. Pineda, J.A., Allison, B.Z., Vankov, A., 2000. The effects of self-movement, observation, and imagination on mu rhythms and readiness potentials (RP's): toward a braincomputer interface (BCI). IEEE Trans. Rehabil. Eng. 8 (2), 219–222. Pineda, J.A., Hecht, E., 2009. Mirroring and mu rhythm involvement in social cognition: are there dissociable subcomponents of theory of mind? Biol. Psychol. 80 (3), 306–314. Pineda, J.A., Oberman, L.M., 2006. What goads cigarette smokers to smoke? Neural adaptation and the mirror neuron system. Brain Res. 1121 (1), 128–135.

Prinz, W., 2005. An ideomotor approach to imitation. In: H. S., C. N. (Eds.), Perspective on imitation: from neuroscience to social science. MIT Press, Cambridge, MA, pp. 141–156. Raymaekers, R., Wiersema, J.R., Roeyers, H., 2009. EEG study of the mirror neuron system in children with high functioning autism. Brain Res. 1304. Rizzolatti, G., Fadiga, L., Gallese, V., Fogassi, L., 1996. Premotor cortex and the recognition of motor actions. Brain Res. Cogn. Brain Res. 3 (2), 131–141. Rizzolatti, G., Fogassi, L., Gallese, V., 2001. Neurophysiological mechanisms underlying the understanding and imitation of action. Nat. Rev. Neurosci. 2 (9), 661–670. Rizzolatti, G., Sinigaglia, C., 2010. The functional role of the parieto-frontal mirror circuit: interpretations and misinterpretations. Nat. Rev. Neurosci. 11 (4), 264–274. Salmelin, R., Hari, R., 1994. Characterization of spontaneous MEG rhythms in healthy adults. Electroencephalogr. Clin. Neurophysiol. 91, 237–248. Sauseng, P., Klimesch, W., 2008. What does phase information of oscillatory brain activity tell us about cognitive processes? Neurosci. Biobehav. Rev. 32 (5), 1001–1013. Schnitzler, A., Salenius, S., Salmelin, R., Jousmaki, V., Hari, R., 1997. Involvement of primary motor cortex in motor imagery: a neuromagnetic study. NeuroImage 6 (3), 201–208. Schulte-Rüther, M., Markowitsch, H.J., Shah, N.J., Finka, G.R., Piefke, M., 2008. Gender differences in brain networks supporting empathy. NeuroImage 42, 393–403. Shimada, S., Abe, R., 2010. Outcome and view of the player modulate motor area activity during observation of a competitive game. Neuropsychologia 48, 1930–1934. Singer, T., Seymour, B., O'Doherty, J., Kaube, H., Dolan, R.J., Frith, C.D., 2004. Empathy for pain involves the affective but not sensory components of pain. Science 303 (5661), 1157–1162. Wicker, B., Keysers, C., Plailly, J., Royet, J.P., Gallese, V., Rizzolatti, G., 2003. Both of us disgusted in My insula: the common neural basis of seeing and feeling disgust. Neuron 40 (3), 655–664. Zion-Golumbic, E., Kutas, M., Bentin, S., 2010. Neural dynamics associated with semantic and episodic memory for faces: evidence from multiple frequency bands. J. Cogn. Neurosci. 22 (2), 263–277.