Validation of joint angle measurements: Comparison of a novel low cost marker-less system with an industry standard three-dimensional marker-based system
Introduction/Background Human motion tracking is widely used for assessment of movement dysfunction in orthopaedic patients. Currently, most clinical motion analysis centres use marker based three-dimensional systems as they are deemed to be the most accurate method. However, due to space, costs and logistics they are not available in many clinical settings. This study compared joint angles measured in functional tests using the novel low-cost Microsoft Kinect Perfect-Phorm marker-less system with the established marker based Nexus VICON system.
Material and Method Ten unimpaired volunteers were fitted with 16 retroreflective markers in the positions specified by the Plug-in-Gait (PiG) lower limb marker set. Participants were asked to make three movements: standing, a squat action and a hip abduction action. The two measurement systems were used simultaneously to measure changes in joint angle. The collection of data from each participant was synchronised manually between each system. Average values for left and right knee flexion (for the squat task) and left hip abduction (for the hip abduction task) were obtained from each system as the average across the steady state, and the relative change between the standing and task sections used as the basis of comparisons of system performance. Results When measuring right and left knee flexion, the average difference between the VICON and Perfect-Phorm measurement was 13.2%, with a SD of 19.6. Both overestimation and underestimation of the joint angle was recorded in different participants. Although the average percentage difference during hip abduction tests was lower at -3.9%, the range of error was far greater (SD=75).
Conclusion From this, it can be concluded that the level of accuracy presented in the new low cost Perfect-Phorm marker-less system is not yet suitable for clinical assessments. However, for general tests of performance, and for tracking cases where absolute accuracy is less critical, future versions of this software may have a place.