Abstract
Subjects were asked to ‘point’ to two targets with their torso, by standing in place and rotating about their vertical z-axis. The targets, which were extinguished at movement onset, were separated by 60 deg of azimuth at a distance of 55 cm from the torso rotation axis. After turn onset, the floor bearing both the subject and the extinguished targets was servo-rotated in proportion to rotation velocity. This altered the relation between the visually specified movement amplitude and the force output required to make accurate ‘pointing’ movements. The gain of the floor velocity relative to the torso rotation velocity was incremented to either + or −0.5 in two separate sessions. A gain of +0.5 meant a 50% increase in the spatial size of the turn required to point to the targets. Subjects received visual feedback and a chance to correct errors only after the completion of each movement. During the final block of positive gain exposure, torso rotation amplitude and velocity relative to inertial space were 32% and 29% greater than during baseline. In the last block of negative feedback trials, torso inertial displacement and velocity were 50% and 54% less than in the baseline period. Following exposure to floor rotation with positive feedback and return to normal stationary floor conditions, torso rotations initially displayed increased movement amplitude; movement amplitudes decreased as an aftereffect of negative feedback. This pattern of aftereffects indicates that adaptation involved compensatory increases and decreases, respectively, in the foot-to-torso torques generated during rotation. No subject was aware of the large physical changes in head and body rotation relative to inertial space that occurred during exposure, which stimulated their semicircular canals at levels well above normal thresholds for detecting velocity differences. This paradigm demonstrates that rapid motor adaptation of torso rotation control to altered dynamic forces is possible.
NIH F32 NS11154-03 & NIH R01 AR48546-01