No abstract

It is assumed that imagining oneself from a first-person perspective (1PP) is more embodied than a third-person perspective (3PP). Therefore, 1PP imagery should lead to more activity in motor and motor-related structures, and the postural configuration of one's own body should be particularly relevant in 1PP simulation. The present study investigated whether proprioceptive information on hand position is integrated similarly in 1PP and 3PP imagery of hand movements. During functional magnetic resonance imaging (fMRI) scanning, 20 right-handed female college students watched video sequences of different hand movements with their right hand in a compatible versus incompatible posture and subsequently performed 1PP or 3PP imagery of the movement. Results showed stronger activation in left hemisphere motor and motor-related structures, especially the inferior parietal lobe, on 1PP compared with 3PP trials. Activation in the left inferior parietal lobe (parietal operculum, SII) and the insula was stronger in 1PP trials with compatible compared with incompatible posture. Thus, proprioceptive information on actual body posture is more relevant for 1PP imagery processes. Results support the embodied nature of 1PP imagery and indicate possible applications in athletic training or rehabilitation.

The simulation concept suggested by Jeannerod (Neuroimage 14:S103-S109, 2001) defines the S-states of action observation and mental simulation of action as action-related mental states lacking overt execution. Within this framework, similarities and neural overlap between S-states and overt execution are interpreted as providing the common basis for the motor representations implemented within the motor system. The present brain imaging study compared activation overlap and differential activation during mental simulation (motor imagery) with that while observing gymnastic movements. The fMRI conjunction analysis revealed overlapping activation for both S-states in primary motor cortex, premotor cortex, and the supplementary motor area as well as in the intraparietal sulcus, cerebellar hemispheres, and parts of the basal ganglia. A direct contrast between the motor imagery and observation conditions revealed stronger activation for imagery in the posterior insula and the anterior cingulate gyrus. The hippocampus, the superior parietal lobe, and the cerebellar areas were differentially activated in the observation condition. In general, these data corroborate the concept of action-related S-states because of the high overlap in core motor as well as in motor-related areas. We argue that differential activity between S-states relates to task-specific and modal information processing.