Author:

Keywords:

EXTRASTRIATE CORTEX, MOTION PERCEPTION, VISUAL MOTION

Abstract:

A substantial fraction of neurons in ventral premotor area F5 have mirror properties, responding to the observation and execution of grasping actions (Gallese et al Brain 1996). Monkey fMRI studies showed that F5 is activated by observation of grasping actions relative to static and dynamic controls, and that its different sectors can be distinguished: F5a responds to observing a person grasping and to observing an isolated hand grasping, while F5c responds only to the first of these stimuli and F5p little to either of them (Nelissen et al Science 2005). The present experiments were undertaken 1) to validate in the monkey the distinction between observing a person grasping and a hand performing the same action by better controlling for visual differences and 2) to use these two fMRI tests in humans to localize the putative human homologue of F5c. A new isolated hand grasping video was made by cutting out the moving hand from the person acting video. Three monkeys were scanned (Siemens Trio 3T, contrast enhanced) with this new video, the acting person video, and their static control conditions (new paradigm). In humans, 18 subjects were first scanned (Philips Achieva, 3T) with the initial paradigm of Nelissen et al 2005, comparing the two types of actions to static and scrambled controls (experiment 1). Subsequently, 5 subjects were scanned both with the old and the new paradigm (experiment 2). In the monkey the results obtained with the new paradigm confirmed the results of Nelissen et al 2005: while F5a was activated, relative to the static control, both by observing a person grasping an object and by observing an isolated hand grasping the same object, F5c was differentially activated only by observing the person acting, and F5P showed little activation. Thus the interaction between the factor actor (present or not) and type of video (action or control) is a marker for F5c in both paradigms, provided one requires a visual response in the person acting condition. In humans, observation of an isolated hand grasping activated the posterior bank of the precentral sulcus between the inferior and superior frontal sulci, bilaterally (experiment 1). The interaction between the factor actor and type of video, however, was significant in the right hemisphere in the middle of the region activated by observing the isolated hand grasping (40, -4, 44). This site was active when observing the person acting was compared to fixation. This result was confirmed in experiment 2 in a similar location (52, -4, 40) and was obtained not only with the old, but also the new paradigm. These results suggest that the interaction site in right area 6 is the putative human homologue of monkey F5c, housing mirror neurons.