ALE meta-analysis of action observation and imitation in the human brain
Introduction
The neural bases of action observation and action imitation in the human brain have been a longstanding interest of neuroscientific research. Increasing attention was focused on these functions and their neuronal correlates when “mirror neurons” were identified in the macaque brain using single-cell recordings (Gallese et al., 1996, Fogassi et al., 2005). These neurons are active not only when performing an action but also when observing another subject performing the same action (Gallese et al., 1996). This discovery in the macaque brain raised the question of whether a comparable system also exists in humans (e.g., Rizzolatti et al., 2001). However, since single-cell recordings are rarely feasible in humans, a direct demonstration of mirror properties for individual human neurons has not yet been provided. Consequently, evidence for possible “mirror” areas in humans is predominantly based on the results of functional neuroimaging experiments. Over the last decade, several studies using functional magnetic resonance imaging (fMRI) and positron emission tomography (PET) have investigated different aspects of action processing in the human brain (e.g., Buccino et al., 2004b, Iacoboni et al., 1999) that are conceptually related to “mirror” properties, in particular action observation and imitation.
Investigation into the human action observation network directly relates to the properties of mirror neurons as defined in nonhuman primates. It is assumed that observing actions enables the mirror neuron system to understand the actions themselves as well as the underlying intentions (e.g., Fabbri-Destro and Rizzolatti, 2008, Rizzolatti, 2005, Rizzolatti and Fabbri-Destro, 2008). By understanding the action with one's own motor system, it is possible to infer on the intentions behind a motor act (e.g., Prinz, 2006, Schütz-Bosbach and Prinz, 2007), a mechanism that already has been proposed long before the discovery of mirror neurons (e.g., Viviani and Terzuolo, 1973). Such ability is then seen as a crucial step towards the development of complex interpersonal and social interactions as witnessed in humans but also other primates (Iacoboni, 2009, Rizzolatti and Fabbri-Destro, 2008).
Understanding an action and its intention might also provide an important link between the sole observation of an action and its subsequent imitation by directly copying the observed action (e.g., Fabbri-Destro and Rizzolatti, 2008, Rizzolatti and Craighero, 2004, Rumiati et al., 2005). Furthermore, imitation offers a potential mechanism for learning from the early stages of life. The motor system can learn how specific actions are carried out by imitating them (e.g., Bandura and Wood, 1989, Brass and Heyes, 2005, Iacoboni, 2005), a mechanism that has long been discovered much earlier in human neonates (Meltzoff and Moore, 1977). Furthermore, just like action understanding, imitation processes play an important role during social interactions: people also tend to imitate behaviours of their social partners (either consciously or subconsciously) to adapt to a given social situation (e.g., Bargh et al., 1996, Iacoboni, 2009, Niedenthal et al., 1985, Schilbach et al., 2008a).
Therefore, assessment of the neural substrates of both action observation and action imitation is not only important for understanding action-related processes but also holds further implications for cognitive and social neuroscience. In spite of the considerable number of neuroimaging studies on these action-related topics, the organisation of the respective networks in the human brain and their anatomical correlates are still disputed (Dinstein et al., 2008, Iacoboni, 2005, Iacoboni, 2009, Keysers and Gazzola, 2009). One controversial aspect is the role of Broca's region in action- related processes (Brass and Heyes, 2005, Molenberghs et al., 2009, Molnar-Szakacs et al., 2005, Vogt et al., 2007). Another is the hemispheric dominance of such functions, as arguments have been made for a leading role of either hemisphere as well as for a bilateral distribution (e.g., Iacoboni and Dapretto, 2006). Finally, since observation and imitation are closely related, the question of whether they are sustained by the same neuronal networks or engage different brain areas is still disputed (e.g., Heyes, 2001, Brass and Heyes, 2005, Turella et al., 2009a, Turella et al., 2009b).
One reason for the diverging evidence on the involvement of different brain regions in these networks is the heterogeneity of the experimental approaches, such as paradigms and effectors (e.g., hand/fingers, face, feet), that have been used to delineate the neural correlates of these functions. To identify those areas in the human brain that are consistently implicated in action processing, the results of these different studies should be synopsized in a quantitative, unbiased fashion. Previous summaries of published studies on action observation or imitation have consisted of qualitative reviews of the reported activation sites (e.g., Brass and Heyes, 2005, Fabbri-Destro and Rizzolatti, 2008, Iacoboni, 2005, Iacoboni, 2009, Rizzolatti et al., 2001). However, a promising new approach for identifying the neural substrates of action observation and imitation in humans is the use of coordinate-based meta-analysis. These analyses aim at revealing areas that are consistently activated in a particular class of paradigms (Laird et al., 2005a, Laird et al., 2009, Eickhoff et al., 2009).
The aim of the present study was to provide a quantitative meta-analysis of the current neuroimaging literature to delineate consistently activated cortical regions associated with action observation and imitation. In a first step, the neural correlates of these processes were analysed separately. Additional subanalyses that assessed the effects of potential confounds, such as effectors or instructions, were carried out to evaluate the consistency of the findings. Conjunction and contrast analyses were performed to reveal divergent and convergent areas for action observation and imitation. Using probabilistic cytoarchitectonic maps of cortical areas, activations identified in each analysis were specifically allotted to the most probable brain area.
Section snippets
Data used for the meta-analysis
Functional imaging studies included in the meta-analysis were obtained from the BrainMap database (www.brainmap.org; Fox and Lancaster, 2002, Laird et al., 2005b) and a PubMed literature search (www.pubmed.org, search strings: “mirror neurons”, “imitation”, and “action observation”) on functional magnetic resonance imaging (fMRI) and positron emission tomography (PET) experiments. The literature cited in the obtained papers was also assessed to identify additional neuroimaging studies dealing
Action observation network
Brain regions showing consistent activation across the 104 action observation experiments were observed symmetrically across both hemispheres in frontal areas BA 44/45, lateral dorsal premotor cortex (dPMC, BA 6), supplementary motor area (SMA, BA 6), rostral inferior parietal lobule (IPL, area PFt), primary somatosensory cortex (SI, BA 1/2), superior parietal (SPL, area 7A), intraparietal cortex (IPS, area hIP3), posterior middle temporal gyrus (pMTG) at the transition to visual area V5, and
Discussion
The present study assessed the action observation and imitation networks in the human brain in a meta-analysis of 139 fMRI and PET experiments. Both action observation and imitation experiments were consistently associated with activation in a largely bilateral network of premotor, primary somatosensory, inferior parietal, and intraparietal as well as temporo-occipital areas. Further analysis revealed that this activation pattern is largely independent from possible confounds, such as
Conclusions
In the present quantitative meta-analysis of neuroimaging data, we identified the cortical regions that are consistently implicated in the human observation and imitation networks. Hereby, the findings of 139 individual experiments could, for the first time, objectively be generalized in an unbiased fashion.
It was shown that action observation and imitation are sustained by a bilateral network spanning fusiform, posterior temporal, parietal, and premotor areas including BA 44. These activation
Acknowledgments
This Human Brain Project/Neuroinformatics Research was funded by the National Institute of Biomedical Imaging and Bioengineering, the National Institute of Neurological Disorders and Stroke and the National Institute of Mental Health. Further funding was granted by the Human Brain Project (R01-MH074457-01A1; S.B.E., A.R.L.), the Initiative and Networking Fund of the Helmholtz Association within the Helmholtz Alliance on Systems Biology (Human Brain Model; K.Z., S.B.E.), and the Helmholtz
References (200)
- et al.
Cytoarchitecture of the cerebral cortex—more than localization
NeuroImage
(2007) - et al.
Congruent embodied representations for visually presented actions and linguistic phrases describing actions
Curr. Biol.
(2006) - et al.
Listening to a walking human activates the temporal biological motion area
NeuroImage
(2005) - et al.
Imitation: is cognitive neuroscience solving the correspondence problem?
Trends Cogn. Sci.
(2005) - et al.
Neural circuits underlying imitation learning of hand actions: an event-related fMRI study
Neuron
(2004) - et al.
The mirror neuron system and action recognition
Brain Lang.
(2004) - et al.
Seeing or doing? Influence of visual and motor familiarity in action observation
Curr. Biol.
(2006) - et al.
The human inferior parietal lobule: cytoarchitectonic parcellation and interindividual variability
NeuroImage
(2006) - et al.
Does the end justify the means? A PET exploration of the mechanisms involved in human imitation
NeuroImage
(2002) - et al.
An fMRI study of imitation: action representation and body schema
Neuropsychologia
(2005)