A versatile robotic platform for the design of natural, three-dimensional reaching and grasping tasks in monkeys

Abstract

OBJECTIVE: Translational studies on motor control and neurological disorders require detailed monitoring of sensorimotor components of natural limb movements in relevant animal models. However, available experimental tools do not provide a sufficiently rich repertoire of behavioral signals. Here, we developed a robotic platform that enables the monitoring of kinematics, interaction forces, and neurophysiological signals during user-defined upper limb tasks for monkeys. APPROACH: We configured the platform to position instrumented objects in a three-dimensional workspace and provide an interactive dynamic force-field. MAIN RESULTS: We show the relevance of our platform for fundamental and translational studies with three example applications. First, we study the kinematics of natural grasp in response to variable interaction forces. We then show simultaneous and independent encoding of kinematic and forces in single unit intra-cortical recordings from sensorimotor cortical areas. Lastly, we demonstrate the relevance of our platform to develop clinically relevant brain computer interfaces in a kinematically unconstrained motor task. SIGNIFICANCE: Our versatile control structure does not depend on the specific robotic arm used and allows for the design and implementation of a variety of tasks that can support both fundamental and translational studies of motor control.