Definition of the mirror neuron system
Neuroanatomy of the mirror neuron system motor/imitation :
There are two main neuronal networks that make up the mirror neuron system (Cattaneo & Rizzolati, 2008): one formed by areas of the parietal lobe and the premotor cortex, as well as by the caudal part of the lower frontal rotation; and another formed by the isula and the medial anterior frontal cortex. We will now focus on the first system, which involves learning based on observation and imitation. The anatomical organization of the first system responds to a somatopic hierarchy of the ventral premotor cortex, being the motor acts of the legs located in the dorsal zone; the facial behaviors located ventrally, and the manual ones with an intermediate distribution. The location of proximal motor acts (moving the hand towards a point) are represented dorsally, while the simple act of grasping produces a ventral activity in the premotor cortex. On the other hand, the observation of motor acts produces a differential activation in the parietal cortex, as well.
The observation of transitive acts produces an activation of the intraparietal sulcus, as well as an activation of the parietal convexity adjacent to that area. The observation of intransitive acts -regardless of whether they are symbolic acts or mimic repetition- find a specific activity in the posterior part of the supramarginal gyre, which extends to the angular gyre. Finally, the observation of acts performed with tools specifically activates the most rostral part of the supramarginal turn. The mirror neuron system produces an evocation of the observed motor actor within the premotor cortex itself. This activity is coordinated at the same time in the parietal lobe. It is necessary to differentiate the sequence of observation processes in order to correctly delimit the neuronanatomy of the first exposed system (frontoparietal). In this system, we will talk about observed behaviors that suppose a visuomotor priming for the execution -or not- of an action. Therefore, we will exclude the conception of motorvisual priming, which implies predictions of consequences during the planning of actions. The activity of this frontoparietal system -and this is what is relevant- occurs when the behavior exists -potentially- in the subject’s repertoire. In other words, a human observing a bark does not activate the premotor and parietal areas, since it does not possess that behavioral repertoire in the cortex.
On the other hand, the activity of the system is proportional to the experience of the observer in the behavior he is observing. The functional connectivity of the frontoparietal system of mirror neurons presents a sequence during observation. Originally this sequence originates in the occipital lobe, where the main characteristics of the observed stimuli are recorded. All the information is sent to the integration areas in a series of steps that vary from 20 msec. to 60 msec., in this order: first the upper temporal furrow, then to the lower parietal lobe, after that the information goes to the lower frontal turn and finally to the primary motor cortex.
Iacoboni et al. (1999) propose that the frontal zones that are activated suppose a computation of the goals to be achieved, while the parietal activity corresponds to the activation of the motor representations of the acts that are being observed. However, Iacobini’s group makes a functional differentiation in the neuronal activity of the system, focusing on the opercularis pars of the lower left frontal rotation. For them, the dorsal zone of the pars is activated when the act is observed and when it is imitated; but a ventral activity only occurs when it is imitated. In fact, Iacoboni et al. (2005) functionally analyse the activations described above.
For them, the mirror neuron system is fundamental for learning through imitation. And the activation sequence would be completed as follows: (i) first there is an activation of the upper temporal furrow, where the ventral representations of the observed movements are found. From there (ii) is passed to a codification of action goals, through the frontoparietal system, in which the dorsal prefrontal cortex would be in charge of computing the different aspects of the action, such as the goal itself or the meaning, archiving this information, sending information to the parietal lobe and correcting computations over space.
This efferent information would be sent from the frontoparietal mirror neuron system, through the opercularis pars, to the upper temporal furrow again. At this point, there would be a computation of the fit that exists between the predicted consequences in the planned imitative action, and the visual description of the observed action. In short, the frontoparietal mirror neuron system constitutes a learning system based on feedback.
In fact, what is transferred from the visual areas to the motor areas is not a detailed motor program, but a prototype of the action, an action with meaning that is processed in the opercularis pars of the lower frontal rotation; and that then guides the motor planning according to a precise detailed representation of the observed action, represented in the upper temporal furrow and in the lower parietal lobe. When the observed action is novel, before the execution period, there is an activation of the frontoparietal mirror neuron system, as well as an activation of the AB 46 area and the anterior medial cortex.
This activation translates into an executive control mechanism, probably as part of the Shallice supervisory mechanism on which Baddeley (2000) relies to formulate the working memory mechanism. In our case, such a system could involve a top-down motion planning computation, in which the working memory manages the observed contents and plans the motion based on them, producing a frontoparietal activity that corresponds to the mirror neuron mechanism. The mirror neuron system should not be conceived as a separate neuronal module, but as an intrinsic mechanism underlying most areas related to motor movements. In fact, and as we will see later, the disruption of this system does not cause a selective deficit in focal lesions. Rather, the implication of this system is proven in disorders of the development of the nervous system, and in lesions of the frontal lobe. Let’s look at this last case below.
Dependence and hierarchy
As mentioned above, the mirror neuron system overlaps with other systems, and the control system is no exception, as it suppresses spontaneous imitation behaviors. Frontal injuries cause a series of deficits characterized by the appearance of impulsive behaviors generated by external stimuli. Imitative behaviour is of particular relevance to the mirror neuron system and may be part of the “environmental dependency syndrome”. Normally, the condition arises from a bilateral lesion, although it may also be due to a less frequent unilateral lesion. The observation of the behaviour of others can elicit an activation of the premotor and parietal zones, dependent on the mirror neuron system.
In healthy subjects, this activation does not occur because there is suppression by the frontal lobe. Its deterioration implies a destruction of these mechanisms, transforming potential acts into acts of fact. The ecopraxia constitutes the forced and critical imitation of observed behaviors, normally with the presence of perseverations. Although it usually occurs as a disorder associated with damage to the basal ganglia, it is also produced by a frontal deterioration, which produces a disinhibition of the mirror neuron system.
Functionality of the frontoparietal system of mirror neurons
Imitation and learning
A fundamental task of learning is imitation, which produces the development of some basic skills of social development, especially in the acquisition of gestural identification, postural, and allows the development of understanding of the intentionality of the other.
These neurons are triggered when the subject performs behaviours related to a goal, but especially when he observes these behaviours in others, discriminating between the different components of the action according to whether they are more or less relevant from the intentional point of view; even before objects that are not present. From the foregoing it can be deduced that mirror neurons not only handle contents related to motor or visual patterns, but also abstract, both in terms of the sensory modality of the contingency (a sound with meaning) and in terms of elements of a non-present or abstract nature, which present a relationship, in terms of learning, with intentionality, a reality in which the understanding of the motives of others plays an important role. The integrated motor information presents significant procedural characteristics: processing of movement, of parts of the body, follow-up of the action aimed at a goal of an external subject, etc.
The proximity to frontoparietal systems that support various types of sensorimotor integration suggests that the coding of the action implemented in the mirror neuron system is linked to some form of sensory integration. Imitation is one of the many forms of this type of integration. In such integration, the observing subject makes comparisons between the existing information in the primary areas (visual inputs) and the observed behavior, as explained above.
The literature on imitation behaviors emphasizes that an essential aspect in this area is the differentiation between various forms of imitation or contagion, and true imitation – that is, adding something new to one’s own motor repertoire after observing others performing that action. This differentiation is observed at the neuronal level, distinguishing the interactions between the system of mirror neurons and preparation structures for prefrontal and parietal execution during learning by imitation, and the interaction between the system of mirror neurons and the limbic system during emotional contagion. Probably, as we will discuss later, the mirror neuron system in autism also allows this distinction to be made, with one of the interaction systems being more damaged than the other.
Mirror neurons have individual properties:
They are activated in the imitated action, but also in the action that is being observed even without imitating it. They have two levels of congruence: strict, in which neurons are activated exclusively in substantially identical actions and observations; and approximate congruence, in which they are activated in response to the observation of an action that is not necessarily identical to the action performed, but achieves the same goal. Activation thresholds are defined by the logic of the action, not by the object or the distance of the action. From these properties it can be inferred that they handle abstract contents of the observed actions. But what is the degree of abstraction of this codification? High, as demonstrated in experiments with previous conditions of “concealment”, in which neurons are activated on the basis of an initial situation of presence or absence, discriminating situations.
There is a sensory recognition of sonorous actions (sonorous inputs) in the mirror neuron system. This provides a basis for understanding speech and language as a code that is learned – at least in initial phases – through physical and gestural imitation.
Functional hierarchy of the frontoparietal system in motor processing
As mentioned above, there is a functional hierarchy in the mirror neuron system when the subject observes a motor action in order to learn it. The basic levels of motor processing have been extensively studied. However, the mirror neuron system responds to a hierarchy in which the processing of movements is of high rank, producing computations between the consequences of the action and the goals.
In order to compute such knowledge, the components that present the context of the action must be dissociated: first, the object itself, which is the goal. Existing studies have not been conclusive until relatively recently. However, by means of neuronal suppression techniques such as Magnetic Stimulation, high-end processing has been dissociated from merely kinematic processing. It has been observed that target-object identification is computed in the anterior intraparietal sulcus (Hamilton & Grafton, 2006). Therefore, there is differential processing of objects, even if the action is the same (e.g. gripping). On the other hand, this dissociation also implies the analysis of the expected consequences of the action, which presents a higher hierarchy level than the previous one.
It is very important to bear in mind that goal processing implies the processing of the movements necessary to achieve that goal, but they are aspects with a different level of processing, being the processing of the motor program (not its planning) a lower processing range. Hamilton & Grafton (2007) demonstrated that there is a lateralization of the system that computes the consequences of the action. They found that the consequences of an observed action are processed in the lower frontal rotation and lower right parietal lobe, as well as in the left postcentral groove and in the left intraparietal anterior groove.
Together they have proposed a hierarchical model that is composed as follows: On the one hand, there is a low-level -cognitive- processing that involves the processing of the motor pattern. The processing of the motor pattern takes place in a system that involves both visual and motor analysis of the action. Visual processing would be performed in lateral occipital areas, while kinematic pattern processing is performed in the lower frontal regions.
High-level processing, defined by an analysis of goals, is performed in a system involving two areas of the right hemisphere: the intraparietal lobe and the lower frontal rotation – to a lesser degree. In this goal processing, goal-objects are also processed laterally in the left lower parietal cortex. Is there a neuronal hierarchy when the observed actions are executed? Yes, and the hierarchical range differentiates the complexity of the actions, that is, when the actions are simple, when they are complex -composed by different steps-, as well as when they respond to an intentionality. In this case, the lateralization of neuronal activity does not seem so evident.
There is evidence that the planning of simple acts occurs in the motor and premotor cortex, as well as in the left lower parietal cortex. However, it appears that the lower right parietal lobe is involved in complex behaviors that require several steps, such as the London Towers task (Newman et al., 2003). This area seems important for sending feedback on the consequences of the motor act, and next to the cerebellum it can compute corrections of movement in space or in planning.
During the task of imitating the movement of fingers, an increase in the activity of the rostral posterior parietal cortex and in the inferior frontal rotation is observed, zones close to the Broca area, which suggests the implication of these mirror areas in a mechanism of phylogenetic language acquisition (Iacoboni & Dapretto, 2006).
This theory has been supported by several clues. Firstly, a left lateralization of the mirror neuron system has been demonstrated. On the other hand, the activation of the mirror neuron system in the brain of the macaque allows to extrapolate its zones to ours: the areas in the macaque would coincide with the AB 44 of the human, adjacent to the Broca area. From the theory of semantic expression, which proposes that language is learned in a bottom-up process, and from the motor theory of discourse perception, which proposes that the objective of discourse analysis is facial expressions associated with sounds, rather than sounds, it has been discovered that during discourse perception the motor areas of discourse are activated, which coincide with the mirror neuron system. In addition, it has been discovered that the processing of linguistic material produces motor activation, and that the neuronal activity produced by the processing of linguistic material related to parts of the body and actions, activates the somatotopic zones of the brain related to reading.
Social cognition and mirror neurons
Neuroanatomy of the limbic mirror neuron system
The second mirror system is the emotional one. As we have said before, this system is involved in the adoption of empathic behaviors, but it does not necessarily work separately from the first system, although this point will be addressed later. The mirror neuron system is also located in cortical areas that mediate emotional behavior. Observing the pain of others produces an activation of the cingulate cortex, the amygdala, and the insula. The insula is especially important in the integration of sensory representations, both internal and external. It has an agranular structure and is cytoarchitecturally similar to motor areas.
Therefore, the insula functions as a communication node between the limbic system and the somatopic cortical activation associated with pain, both internal and external, which constitutes the evolutionary basis of empathy. However, this base is not unique. The system of empathy would be framed in such a way: In the first place there must exist a node in this system, which is the amygdala, which is necessary for the emotional activation of the subjects.
Secondly, the zones of expression and emotional regulation. As we have said, the zone of emotional expression based on bodily schemes is composed of two structures: first the ínula, which as we have said is the center of integration of interoceptive information. On the other hand we would find the cingulate cortex, which is subdivided in the following way: in front of the classic division between cognitive/dorsal and emotional/rostral processes of the cingulate cortex (Posner et al, 2007); recently it has been proven that there is a division in terms of emotional expression (interoceptive) in the dorsal anterior cingulate cortex, and a regulatory function of emotions in the rostral anterior cingulate cortex (Etkin et al., 2010), which is congruent with an antero-posterior control system orbiting the front (mainly frontal orbiting-tonsillar orbiting).
Thirdly, the high-level processing node composed of the mirror neuron system. This system consists of the insula and the middle anterior frontal cortex. In fact, the island’s mirror neuron system overlaps with the interceptive emotional expression system. The interaction of this system with emotion varies depending on the complexity of the emotional act.
How does the mirror neuron system work in social cognition?
The mirror neuron system works in two ways when it comes to social cognition. First, it is necessary for prediction and attribution of thought (theory of mind). Secondly, it sets in motion mechanisms of affective recognition and expressiveness. The first prediction system has been explained: the observed acts are computed in a system of frontoparietal mirror neurons, together with the consequences.
This system serves as a predictive model that advances evolutionarily: from simple bottom-up behaviors and processes, the neurological system passes over the years to a top-down system of regulation, in which the observed motor schemes are compared with learning over the years, and its purpose is to establish statistical predictive patterns that minimize error (Kilner et al., 2007). This computation is also hierarchical, in the sense that the operations performed respond to a hierarchical distribution of the theoretical axes of the brain. In this hierarchy, the frontal lobe is in charge of making the computation between the observed behavior and the assumed mental state, and the motor cortex, the parietal and the superior temporal furrow would be in charge of integrating the visual information and the stored motor schemes.
Next we will focus on the second system of empathy, which involves the limbic mirror neuron system (insula, cingulate cortex and frontal lobe).
Mirror neurons and empathy
The role of mirror neurons in empathic behaviors such as the adoption of facial gestures and postures in interactive imitative behaviors is basic along with emotional adoption (limbic system). As mentioned above, mirror neurons compute movements in terms of execution consequence, and goals. This knowledge serves as the basis for social cognition, along with the second system of emotional integration. Empathy is not a univocal process. Although there is evidence that observing the punishment of others produces an activation in the amygdala, the anterior cingulate cortex, and the isula -in addition to the thalamus and cerebellum- (Jackson et al., 2005), the entire process probably depends on a large-scale network, with high processing zones influencing or eliciting emotional responses.
In fact, this could be the role of mirror neurons in empathy. Empathy is supported by a large-scale neural network composed of the mirror neuron system, the limbic system, and the insula, which functions as a connecting node between the two systems. Within this network, mirror neurons provide the simulation of facial expressions and gestures observed in others to low-level processing zones, through the insula, causing activity in those zones. And, finally, producing an emotional state in the observer of the observed behavior. In this way an alternative system of emotions is provided to the subject, based on simulation, which allows social cognition in part. This theory is called “simulation theory” (Gallese & Goldman, 1998; quoted in Frith & Frith, 2006), and proposes that in this way we can understand the emotions we observe: through the internal states that provoke us. Therefore, the most common form of empathizing that exists is to adopt the position of the other, literally, to simulate it internally. And again, when we try to adopt the posture of the person expressing their emotions, we do it facially, which activates the limbic system.
In short, mirror neurons present a sensorimotor base of empathy. When talking about this system of mirror neurons and their relationship with empathy, it is necessary to make a distinction: understanding and simulating emotions is not the only step for social cognition, since we must take into account the stable personality of the person in order to make predictions. And in this aspect, it is interesting to make, again, a distinction: neuronally, is it the same to think about the probable behavior and emotion of a person similar to us, than of a different person? It is not. Thinking about someone who is similar to us usually activates areas of the medial ventral prefrontal cortex, especially AB 18, 9, 57 and 10, while thinking about the likely reactions and characteristics of other active areas of the dorsal prefrontal cortex, AB 9, 45 and 42 (Frith & Frith, 2006).
In fact, there is a functional axis of the medial-lateral brain, in which the most central zones are related to the representation of the self and one’s own emotions, while the lateral regions imply a representation of the external world and the others. This hypothesis about a medial-lateral axis is based on the fact that the medial zones tend to have a greater connection with limbic centers and proprioceptive sensory information, and are therefore more influenced by data, while the lateral zones would be more reflective and dependent on representations of the external world. Amodio & Frith (2006) mention a central node in the processing of social cognition: the middle frontal cortex (AB 10).
Mirror neuron system and motor rehabilitation
Although the function of the mirror neuron system in motor learning has been explained, it is interesting to note its implication in the formation of a motor memory bank. The strongest evidence comes from studies by Stefan et al. (2007), in which the authors show how learning a motor sequence through observation enhances the formation of motor memories as opposed to solitary learning. It has been discovered that learning by observation can mediate long-term neuroplasticity processes in the individual, and that this effect is mediated by the mirror neuron system in the motor cortex.
In a study carried out by Ertelt et al. (2008) two groups of patients with infarction in the middle cerebral artery and parasitic limb underwent two different therapies: one with audiovisual clues, and another without clues. The group that followed the training with audiovisual samples of the exercises showed a greater improvement in the paretic limb than the control group. In addition to the above, mirror therapy has been proposed as an alternative that causes changes in plasticity. In it, the patient practices with his healthy limb in front of a mirror in which it is visualized in a parasagittal mirror. This produces a visual illusion of the paretic limb. The results of the therapy show a generation of cortical plasticity.
Mirror neurons and therapy in autism spectrum disorders (Autism and Asperger’s)
Development and dysfunction
There is indirect evidence of mirror neuron activity from the first year of life for predicting the goals of observed subjects (Falck-Ytter etal., 2006; cited in Iacoboni & Dapretto, 2006). In children under 11 years of age, this evidence, although less robust than in adults (which is logical if we think that the system is not fully mature in terms of connections), shows activation rates of mirror neurons in several parameters (mu rhythm suppression, EEG, infrared spectroscopy, functional magnetic resonance) for imitation activities, social competence and empathy. Therefore, although the degree of involvement of the mirror neuron system in social behaviour is not known with certainty today, it is evident that it plays a central role. Probably one of the keys to establish its importance is its dysfunction in children with autism and other communication disorders. In autism, a deficit of neuronal simulation has been proposed in the modeling of observed behaviours, which prevents a correct “experimental understanding” of the others. This deficit has been neurologically proven in the circuit of mirror neurons, in which there are structural abnormalities in subjects with autism spectrum disorders.
This disorder, for example, prevents the correct identification of emotions in facial gestures as there is no adequate activation of the central circuit. However, autistic subjects are able to identify the non-emotional act -although they do not know the purpose for which it is performed-, which suggests a disruption of the limbic mirror neuron circuit more accentuated than the imitation mechanism. This deficit correlates with the severity of the disorder. The data provided on Asperger confirm a similar but less serious deficit (based on difficulty and a temporary delay in the acquisition of behaviors) involving imitation, suggesting that imitation may play an important role in therapy with this type of subjects. Therapy with subjects with autism spectrum disorders may involve the mirror neuron system. There is empirical evidence that at least part of the disorder involves a deficit in imitation and language production, and that the mirror neuron system is involved in the disorder (Wan et al., 2010).
Music therapy has been shown to improve symptoms. Since the sensorimotor system is involved in language processing, and there is also a modulation of motor activity during language processing, it seems logical to think that a way of activating the mirror neuron system could improve both symptoms. And music produces an activity in the system, which favors its alteration (in a positive sense) and provides neuronal plasticity, since music is an act of motor expressiveness as well, and produces an activation, among others, of the Broca area (AB 44). In fact, this type of therapy combined with singing produces beneficial effects in patients with Broca Aphasia, many of whom are able to say the words integrated in a prosody different from the normal one. Compared to the act of speaking, singing produces a bilateral activation of a frontotemporal network, and part of this network shares neurons with the mirror neuron mechanism. This overlap produces an improvement in auditory-motor coordination schemes, an activation deficit observed in people suffering from this type of communication disorder. As for imitation, the degree of implementation and effectiveness varies in the same way as the experimental data presented above. It seems that children with Asperger’s show a good evolution, especially if the therapy is initially carried out with subjects close to the affected, and much more if it is carried out with videos with the own subject. These data have been verified in the suppression of mu waves in the sensorimotor cortex, part of the mirror neuron system.
- Cattaneo, L., & Rizzolatti, G. (2009). The mirror neuron system. Archives of Neurology, 66(5), 557-560.
- Ertelt, D., Small, S., Solodkin, A., Dettmers, C., McNamara, A., Binkofski, F., & Buccino, G. (2007). Action observation has a positive impact on rehabilitation of motor deficits after stroke. NeuroImage, 36 Suppl 2, T164-173.
- Frith, C. D., & Frith, U. (2006). The neural basis of mentalizing. Neuron, 50(4), 531-534.
- Grafton, S. T., & Hamilton, A. F. D. C. (2007). Evidence for a distributed hierarchy of action representation in the brain. Human Movement Science, 26(4), 590-616.
- Hickok, G. (2010). The Role of Mirror Neurons in Speech and Language Processing. Brain and language, 112(1), 1.
- Iacoboni, M. (2009). Imitation, empathy, and mirror neurons. Annual Review of Psychology, 60, 653-670.
- Iacoboni, M., & Dapretto, M. (2006). The mirror neuron system and the consequences of its dysfunction. Nature Reviews. Neuroscience, 7(12), 942-951.
- Jackson, P. L., Meltzoff, A. N., & Decety, J. (2005). How do we perceive the pain of others? A window into the neural processes involved in empathy. NeuroImage, 24(3), 771-779.
- Kemmerer, D., & Castillo, J. G. (2010). THE TWO-LEVEL THEORY OF VERB MEANING: AN APPROACH TO INTEGRATING THE SEMANTICS OF ACTION WITH THE MIRROR NEURON SYSTEM. Brain and language, 112(1), 54-76.
- Kilner, J. M., Friston, K. J., & Frith, C. D. (2007). Predictive coding: an account of the mirror neuron system. Cognitive Processing, 8(3), 159-166.
- Le Bel, R. M., Pineda, J. A., & Sharma, A. (2009). Motor-auditory-visual integration: The role of the human mirror neuron system in communication and communication disorders. Journal of Communication Disorders, 42(4), 299-304.
- Oberman, L. M., & Ramachandran, V. S. (2007). The simulating social mind: the role of the mirror neuron system and simulation in the social and communicative deficits of autism spectrum disorders. Psychological Bulletin, 133(2), 310-327.
- Oztop, E., Kawato, M., & Arbib, M. (2006). Mirror neurons and imitation: a computationally guided review. Neural Networks: The Official Journal of the International Neural Network Society, 19(3), 254-271.
- Rizzolatti, G., Fabbri-Destro, M., & Cattaneo, L. (2009). Mirror neurons and their clinical relevance. Nature Clinical Practice. Neurology, 5(1), 24-34.
- Vogt, S., & Thomaschke, R. (2007). From visuo-motor interactions to imitation learning: behavioural and brain imaging studies. Journal of Sports Sciences, 25(5), 497-517.
- Wan, C. Y., Demaine, K., Zipse, L., Norton, A., & Schlaug, G. (2010). From music making to speaking: engaging the mirror neuron system in autism. Brain Research Bulletin, 82(3- 4), 161-168.