The effect of coordinate system variation on in vivo patellofemoral kinematic measures.

Abstract

The use of different coordinate system definitions for the patella leads to difficulties in comparing kinematic results between studies. The purpose of this work was to establish the effect of using a range of coordinate system definitions to quantify patellar kinematics. Additionally, intra- and inter-investigator repeatabilities of the digitization of anatomic landmarks on the patella were determined. Four different patellar coordinate system definitions were applied using digitisations in two and three dimensions and a single femoral coordinate system was used for comparison. Intra-investigator variability was established by having one investigator digitize the patellar landmarks of three subjects on five separate occasions. Inter-investigator variability was quantified by having five participants digitize the same landmarks on the same three subjects. Patellofemoral kinematics were quantified for ten subjects, at six angles of tibiofemoral flexion, using MRI. As a result of changes in the patellar coordinate system, differences of up to 11.5° in flexion, 5.0° in spin, and 27.3° in tilt were observed in the resultant rotations for the same motion, illustrating the importance of standardizing the coordinate system definition. To minimize errors due to variability while still maintaining physiologically sensible kinematic angles, a coordinate system based upon an intermediate flexion axis between the most medial and lateral points on the patella, and a superiorly-directed long axis located between the most proximal and distal points on the patella, with an origin at the centre of the most proximal, distal, medial, and lateral points on the patella is recommended. The recommended anatomic coordinate frame may be employed in the calculation of dynamic in vivo patellar kinematics when used in combination with any method that reliably quantifies patellar motion.