Novel Robotic Control Methods That Account for System Compliance Decrease the Errors in Ligament Tensions Computed Using the Superposition Method

Abstract

Abstract. Superposition testing is a method to quantify in situ ligament tension by measuring the change in joint loads before and after ligament sectioning. Despite its widespread use, the traditional robot control method used in superposition testing may introduce errors because it does not account for system compliance when prescribing joint kinematics. Using the tibiofemoral joint as a model system, our objectives were to quantify the errors in superposition-computed tensions using both the traditional robot control method and a novel sensor fusion control method that accounts for system compliance. Using our industrial robotic testing system, we performed superposition testing to quantify lateral collateral ligament (LCL) tension in five cadaveric knees during prescribed varus and external rotation loading using both robot control and sensor fusion control. We computed the errors between superposition-computed tensions and reference standard ligament tensions measured by an in-series load cell. Compared to robot control, full sensor fusion control significantly decreased the errors in superposition-computed tensions in the lateral collateral ligament at peak tension under 15 N·m applied varus (from –92±30 N to –27±21 N) and 5 N·m applied external rotation (from –27±19 N to –10±9 N) loading by decreasing errors in joint kinematics and bone positions. Further work is needed to determine whether similar errors occur across different joints, ligaments, experimental setups, or applied loads. In applications where errors are unacceptable for a particular context of use, control methods that account for system compliance show promise for improving the accuracy of superposition-computed ligament tensions.