A 5-Digit Manipulandum System to Measure and Train Finger Coordination During Object Manipulation

Knowledge of how high-dimensional finger forces coordinate to achieve low-dimensional object manipulation is of fundamental interest in the field of sensorimotor neuroscience. We developed a novel 5-digit manipulandum integrating force sensors, actuators, and real-time software to measure and train finger force coordination within the conceptual framework of body-machine interfaces. The device enables personalized mapping between fingertip forces and two-dimensional cursor motions through a calibration process based on principal component analysis. In a proof-of-concept study, eight participants learned to maneuver a cursor across a planar target grid within 1 h of training. Significant improvements in movement time (p = 0.002) and path efficiency (p = 0.002) coincided with an increase in the planarity of digit-force coordination patterns (p = 0.001). The improvements in performance and the low-dimensional organization of digit-force coordination were consistent with short-term retention in the sense that performance values after a 6 h delay were comparable to those observed immediately after practice. The results demonstrate the system’s potential as a research tool for studying hand motor control.