Dynamic characterization of upper limb discoordination following hemiparetic stroke

The presence of discoordination following stroke has been described qualitatively in the clinic and quantitatively in static conditions. Under dynamic conditions, aspects of arm movements have been explored in two conditions: supported and unsupported against gravity. While these methods have provided insight into movement patterns following stroke, the picture is incomplete and requires a device that can measure joint torques during dynamic arm motions and provide levels of partial support to the arm during movement. Here, we describe the arm coordination training 3-D (ACT^3D) robotic system and how its unique characteristics can provide additional insight into the dynamic expression of these synergies. Implications and clinical applications are briefly discussed.