Understanding actions of others: the electrodynamics of the left and right hemispheres. A high-density EEG neuroimaging study

Abstract

Background: When we observe an individual performing a motor act (e.g. grasping a cup) we get two types of information on the basis of how the motor act is done and the context: what the agent is doing (i.e. grasping) and the intention underlying it (i.e. grasping for drinking). Here we examined the temporal dynamics of the brain activations that follow the observation of a motor act and underlie the observer's capacity to understand what the agent is doing and why.

Methodology/principal findings: Volunteers were presented with two-frame video-clips. The first frame (T0) showed an object with or without context; the second frame (T1) showed a hand interacting with the object. The volunteers were instructed to understand the intention of the observed actions while their brain activity was recorded with a high-density 128-channel EEG system. Visual event-related potentials (VEPs) were recorded time-locked with the frame showing the hand-object interaction (T1). The data were analyzed by using electrical neuroimaging, which combines a cluster analysis performed on the group-averaged VEPs with the localization of the cortical sources that give rise to different spatio-temporal states of the global electrical field. Electrical neuroimaging results revealed four major steps: 1) bilateral posterior cortical activations; 2) a strong activation of the left posterior temporal and inferior parietal cortices with almost a complete disappearance of activations in the right hemisphere; 3) a significant increase of the activations of the right temporo-parietal region with simultaneously co-active left hemispheric sources, and 4) a significant global decrease of cortical activity accompanied by the appearance of activation of the orbito-frontal cortex.

Conclusions/significance: We conclude that the early striking left hemisphere involvement is due to the activation of a lateralized action-observation/action execution network. The activation of this lateralized network mediates the understanding of the goal of object-directed motor acts (mirror mechanism). The successive right hemisphere activation indicates that this hemisphere plays an important role in understanding the intention of others.

Figure 1. Stimuli and experimental design.

A. Exemplars of the used stimuli. 1: No context condition. Participants observed two pictures in sequence. The first showed an object (e.g., a cup) without any context, the second a hand interacting with that object. Three types of hand-object interactions were presented: a hand grasping the object as for using it (Ug); a hand grasping an object as for moving it (Tg); a hand touching an object without any obvious purpose (Sc). 2: “Context” condition. As in the previous condition, participants saw two pictures in sequence. The first showed an object embedded in a context (upper row). The second one showed a hand grasping that object. The context allowed the observer to decide whether the agent's intention was to use the object or to move it (U and T, middle row). A simple contact of the hand with the object was also presented in both the contexts (Usc: use context, simple contact; Tsc; transport context, simple contact; upper right, lower row). B. Procedure. Each trial consisted of the following sequence: fixation cross; object alone or within a context hand-object interaction; fixation cross. The sequence shown in figure illustrates the use grip (Ug) in No context condition.

Figure 2. No Context condition.

Visual event-related potentials (VEPs) and brain microstates. A. Group-averaged VEPs elicited by the presentation of use grip (Ug), transport grip (Tg) and simple contact (Sc) stimuli. Stimulus exemplars of the three classes of stimuli are shown on the left side of the panel. The topographic cluster analysis identified five distinct microstates (colored bars) in the 500 ms following stimulus presentation. Two microstates (blue and green bars, respectively), although present following presentation of all three types of stimuli, had different duration depending upon the stimulus type. B. Segmentation maps of the two microstates (Microstate blue frame, Microstate green frame) that showed different duration according to the presented stimulus type. The maps are plotted with the nasion upward and right ear on the right side (scale indicated). Blue areas depict negative potentials and red areas depict positive potentials. C. The statistical significance of Microstate 3 and 4 was tested by means of a fitting procedure based on the spatial correlation between microstates obtained from the group-averaged VEPs and the single-subject VEP data. Blue columns: Microstate 3. Note the more prolonged activity for use and transport grip (Ug and Tg,) than for simple contact actions (Sc). Green columns: Microstate 4. Note the shorter activity in response to use grip (Ug) than in response to the other two stimuli. Error bars indicate standard deviation. Asterisks indicate significant differences (**, P<0.01) between conditions for a given microstate. D. Localization of the intracranial brain generators as estimated with LORETA inverse solution. Twelve transaxial brain sections are presented. Their Talairach z values, from left to right and from top to bottom, are: 72, 64, 49, 42, 31, 22, 16, 7, −6, −10, −32, −38. Group-averaged source estimations were calculated over each time interval and all conditions. The localizations are framed with the same color code as the corresponding microstates in A. E. Segmentation maps of the all microstates. Conventions as in B.

Figure 3. Context condition.

Visual event-related potentials (VEPs) and brain microstates. A Group-averaged VEPs elicited by the presentation of hand grasping stimuli in use (U) and transport (T) context, and by the presentation of simple contact stimuli in the same two contexts (Usc and Tsc). Stimulus exemplars of the four classes of stimuli are shown on the left side of the A panel. The topographic cluster analysis identified six distinct microstates (colored bars) in the 500 ms following stimulus presentation. Two microstates (blue and green bars, respectively), although present following presentation of all three types of stimuli, had different duration depending upon stimulus type. The last microstate (Microstate 6, yellow bar) was identical for T, Usc and Tsc. It was markedly different for U (see text). B. Segmentation maps of the two microstates (Microstates 4, blue; Microstate 5, green, upper part of the column) that showed different duration according to the presented stimulus type. The lower panel of the column shows segmentation map of Microstate 6 (pink) that is specific for the case of U class of stimuli. C. The statistical significance of Microstate 4 and 5 was tested by means of a fitting procedure based on the spatial correlation between microstates obtained from the group-averaged VEPs and the single-subject VEP data. Blue columns: Microstate 4. Note the prolonged activation for grasping actions (U and T) with respect to those for simple contact actions (Usc and Tsc). Green columns: Microstate 5. Note the prolonged responses for Usc and Tsc than for U and T. Error bars indicate standard deviation. Asterisks indicate significant differences (*, p<0.05; **, p<0.01) between conditions for a given map observation. D. Localization of the intracranial brain generators as estimated with LORETA inverse solution. Twelve brain transverse sections are presented (z coordinates as in Figure 2). The localizations are framed with the same color as the corresponding microstates in A. Group-averaged sources estimations were calculated over each time interval and all conditions. E. Segmentation maps of the all microstates. Conventions as in B.