Force level independent representations of predictive grip force-load force coupling: a PET activation study.

The existence of forward internal models is a fundamental principle in theories of predictive motor control. There are indications that internal models are represented in the cerebellum. So far, no conclusive data exist on automated procedures involving predictive motor behavior. In particular, it is unknown whether single or multiple task-specific internal models handle the broad range of behavioral situations in which they occur. Using H2(15)O PET in eight subjects, we examined predictive motor control in an automated grip force-load force coupling task at three differing load force levels. In the experimental condition, subjects pulled a grasped object against an isometric resistance while simultaneously producing anticipatory grip forces. There were three control conditions (pull force isolated; grip force isolated; motor rest). A 2 x 2 factorial design was chosen to reveal the interaction effect of grip force-pull force coupling. The factors were pull force (with/without) and grip force (with/without). Grip and load forces were well matched between experimental and control conditions. Conjunction inference and interaction analyses identified force coupling related activity in the ipsilateral posterior cerebellum that was independent of force levels. Interaction effects were also identified in the anterior cingulate and frontal association regions, the right caudate nucleus, and the left lingual gyrus. These data demonstrate the existence of modular representations for predictive force coupling, with the ipsilateral cerebellum playing a major role. Moreover, the data implicate that the representations for predictive force control are applicable to a range of different environmental affordances.     «

The existence of forward internal models is a fundamental principle in theories of predictive motor control. There are indications that internal models are represented in the cerebellum. So far, no conclusive data exist on automated procedures involving predictive motor behavior. In particular, it is unknown whether single or multiple task-specific internal models handle the broad range of behavioral situations in which they occur. Using H2(15)O PET in eight subjects, we examined predictive moto...     »