Final Report Summary - COGSYSTEMS (Understanding actions and intentions of others)
The present project had four main Objectives: a) Description of the functional organization of monkey area F5, with particular emphasis on the properties of mirror neurons; b) Assessment of the anatomical and functional properties of the cortical network underlying the organization of intentional actions; c) Neural basis of action coding in humans; d) Neurophysiological studies of autism.
a) Area F5. The functional organization of monkey area F5 was studied using new linear multielectrodes probes. This technique allowed us to investigate simultaneously the properties of several single neurons during action execution and observation. The main results of these studies can be summarized as follows. a) In addition to canonical and mirror neurons, F5 contains a further class of neurons that respond both during object and action observation (canonical-mirror neurons). These neurons appear to encode the impending action of the observed agent; b) The classical distinction between areas containing mirror neurons and areas containing canonical neuron is incorrect. Both area F5c and area F5p house both mirror and canonical neurons, although in slightly different proportions; c) Location in space is a critical factor for most F5 neurons. Canonical neurons respond almost exclusively to stimuli presented in the peripersonal space, while mirror neurons may prefer either peripersonal or extrapersonal space; d) “Inaction” neurons were discovered. These neurons discharge both when the monkey is instructed to perform an action and when it is instructed to refrain from acting. There are also “mirror inaction” neurons. These neurons fire when the observed agent is going to refrain from acting. Thus, action representation is also present when the action is “prohibited” .
b) Cortical network underlying the organization of intentional actions. We traced the anatomical connections between the prefrontal areas 46 and 12 and the parieto-frontal network involved in the organization of grasping actions. Subsequently the functional properties of these areas were studied. It was found that in the ventrolateral prefrontal cortex there are neurons that respond selectively to the observation of graspable objects that are identical pragmatically (i.e. identical small spheres), but differ for their semantic properties (i.e. sphere that are edible or inedible). This specific type of canonical neurons discharges during motor execution of grasping-to-eat or grasping-to-place action in a way that is coherent with the semantically encoded objects. The mirror properties of these neurons are under investigation. Finally, we found neurons in the prefrontal cortex (areas 46 and 12) that code the monkey intention to grasp-to-eat or grasp-to-place at an abstract level, that is without contextual sensory information enabling the monkey to make a decision on what action to perform.
c) Neural basis of action coding in humans. The aim of this objective was to test the hypothesis that the mirror mechanism is a general principle of cortical organization in humans. To this purpose we studied, using fMRI: a) the cortical encoding of classes of hand actions differing in their motor goal, b) the areas involved in tool use observation, c) the neural substrate of vitality forms. More recently we approached this issue by recording intra-cortically gamma activity in epileptic patients from regions not affected by the epileptic focus. The results showed that the observer recognize all the studied actions done by others using their own motor repertoire. This is true for actions in which crucial is the understanding of what the other individual is doing (e.g. rubbing the skin of another individual), but also for action in which crucial is the understanding of how the other individual perform the action (vitality forms). In this last case the area crucially involved is the dorso-central part of the insula. Intracortical recordings supported the validity of the conclusion about mirror mechanism as a fundamental cortical mechanism. In a recent study in which this technique was used we showed that the same part of the cingulate cortex which, when excited, trigger laughter is also active when a video-clip showing a person laughing is presented to the patient.
d) Neurophysiological studies of autism. Some years ago we proposed that the basic deficit in autism is a deficit in the development of motor system. This deficit has as a secondary consequence the lack of development of the mirror mechanism in many brain areas and, therefore, a deficit in communicative and social capacities. Two studies were carried out in this project supported this view. In the first we studied the performance of children with autism, their siblings, and typically developing children using the Florida Apraxia Battery. Children with autism showed the lowest performance in all sections of the test. They were, however, particularly impaired during imitation of meaningless gestures and when requested to pantomime actions on imitation or on verbal command. Pantomimes were executed in a rigid stereotyped way. Most importantly, a correlation was found between performance in pantomime actions and the severity of autistic deficits. We conclude that the presence of a rigid internal model prevents the execution of an exact copy of the observed pantomime actions and that the deficit in imitation of meaningless gestures is most likely due to a deficit in the mechanisms responsible for visuomotor transformations.
The second study examined the capacity of children with autism to understand the vitality forms of actions done by others. The results showed that, unlike typically developing individuals, individuals with autism reveal severe deficits in recognizing vitality forms, and their capacity to appraise them does not improve with age. Deficit in vitality form recognition appears, therefore, to be an important trait marker of autism.
a) Area F5. The functional organization of monkey area F5 was studied using new linear multielectrodes probes. This technique allowed us to investigate simultaneously the properties of several single neurons during action execution and observation. The main results of these studies can be summarized as follows. a) In addition to canonical and mirror neurons, F5 contains a further class of neurons that respond both during object and action observation (canonical-mirror neurons). These neurons appear to encode the impending action of the observed agent; b) The classical distinction between areas containing mirror neurons and areas containing canonical neuron is incorrect. Both area F5c and area F5p house both mirror and canonical neurons, although in slightly different proportions; c) Location in space is a critical factor for most F5 neurons. Canonical neurons respond almost exclusively to stimuli presented in the peripersonal space, while mirror neurons may prefer either peripersonal or extrapersonal space; d) “Inaction” neurons were discovered. These neurons discharge both when the monkey is instructed to perform an action and when it is instructed to refrain from acting. There are also “mirror inaction” neurons. These neurons fire when the observed agent is going to refrain from acting. Thus, action representation is also present when the action is “prohibited” .
b) Cortical network underlying the organization of intentional actions. We traced the anatomical connections between the prefrontal areas 46 and 12 and the parieto-frontal network involved in the organization of grasping actions. Subsequently the functional properties of these areas were studied. It was found that in the ventrolateral prefrontal cortex there are neurons that respond selectively to the observation of graspable objects that are identical pragmatically (i.e. identical small spheres), but differ for their semantic properties (i.e. sphere that are edible or inedible). This specific type of canonical neurons discharges during motor execution of grasping-to-eat or grasping-to-place action in a way that is coherent with the semantically encoded objects. The mirror properties of these neurons are under investigation. Finally, we found neurons in the prefrontal cortex (areas 46 and 12) that code the monkey intention to grasp-to-eat or grasp-to-place at an abstract level, that is without contextual sensory information enabling the monkey to make a decision on what action to perform.
c) Neural basis of action coding in humans. The aim of this objective was to test the hypothesis that the mirror mechanism is a general principle of cortical organization in humans. To this purpose we studied, using fMRI: a) the cortical encoding of classes of hand actions differing in their motor goal, b) the areas involved in tool use observation, c) the neural substrate of vitality forms. More recently we approached this issue by recording intra-cortically gamma activity in epileptic patients from regions not affected by the epileptic focus. The results showed that the observer recognize all the studied actions done by others using their own motor repertoire. This is true for actions in which crucial is the understanding of what the other individual is doing (e.g. rubbing the skin of another individual), but also for action in which crucial is the understanding of how the other individual perform the action (vitality forms). In this last case the area crucially involved is the dorso-central part of the insula. Intracortical recordings supported the validity of the conclusion about mirror mechanism as a fundamental cortical mechanism. In a recent study in which this technique was used we showed that the same part of the cingulate cortex which, when excited, trigger laughter is also active when a video-clip showing a person laughing is presented to the patient.
d) Neurophysiological studies of autism. Some years ago we proposed that the basic deficit in autism is a deficit in the development of motor system. This deficit has as a secondary consequence the lack of development of the mirror mechanism in many brain areas and, therefore, a deficit in communicative and social capacities. Two studies were carried out in this project supported this view. In the first we studied the performance of children with autism, their siblings, and typically developing children using the Florida Apraxia Battery. Children with autism showed the lowest performance in all sections of the test. They were, however, particularly impaired during imitation of meaningless gestures and when requested to pantomime actions on imitation or on verbal command. Pantomimes were executed in a rigid stereotyped way. Most importantly, a correlation was found between performance in pantomime actions and the severity of autistic deficits. We conclude that the presence of a rigid internal model prevents the execution of an exact copy of the observed pantomime actions and that the deficit in imitation of meaningless gestures is most likely due to a deficit in the mechanisms responsible for visuomotor transformations.
The second study examined the capacity of children with autism to understand the vitality forms of actions done by others. The results showed that, unlike typically developing individuals, individuals with autism reveal severe deficits in recognizing vitality forms, and their capacity to appraise them does not improve with age. Deficit in vitality form recognition appears, therefore, to be an important trait marker of autism.