Non-invasive measurement of the patellofemoral movements during knee extension-flexion : a validation study

Background: This study compared the difference between patellofemoral kinematics derived simultaneously from patellar bone pin and skin sensors during full range of tibiofemoral joint movement. Methods: Movements at the tibiofemoral and patellofemoral joints during passive full extension–flexion of the knees in three un-embalmed human cadavers were studied with four electromagnetic tracking sensors at a sampling rate of 30 Hz. A total of four sensors were attached on distal femur, proximal tibia, the surface of a tailor-made patella mold and at the tip of a plastic bone pin planted in the patellar body through a window on the mold. Paired-sample Wilcoxon signed rank test was used to compare peak motions computed from different sensors. The correlation of the movement–time curves derived from different sensors was tested by coefficient of multiple correlations (CMC) in different sections of tibiofemoral joint range. Results: Peak motions detected by skin sensor for patellar lateral tilt (p= 0.045), distal translation (p= 0.021), lateral shift (p= 0.032), and anterior–posterior shift of patella (p= 0.03 and 0.01 respectively) were higher than that by the bone pin sensor. The overall CMC values for anterior–posterior translation and medial–lateral shift were lower than movements in other planes of movement. The CMC values in initial range were higher than that in the middle and end range in all planes of movement. Conclusions: Patellofemoral kinematics derived from skin sensors may not be representative of the underlying patellar motion. Kinematics reported from the skin sensors should be carefully interpreted.