Accuracy of a contour-based biplane fluoroscopy technique for tracking knee joint kinematics of different speeds

• Published in: Journal of biomechanics 2012-11, Vol.45 (16), p.2935-2938
• Main Authors: Giphart, J. Erik, Zirker, Christopher A, Myers, Casey A, Pennington, W. Wesley, LaPrade, Robert F
• Format: Article
• Language: English
• Subjects: Accuracy; Applied sciences; Bias; Biological and medical sciences; Biomechanical Phenomena; Biplane fluoroscopy; Biplanes; Bones; Degrees of freedom; Exact sciences and technology; Femur - physiology; Fluoroscopy; Fluoroscopy - methods; Fundamental and applied biological sciences. Psychology; Humans; Hypotheses; Image processing; Information, signal and communications theory; Kinematics; Knee; Knee Joint - physiology; Knee kinematics; Knees; Measurement techniques; Models, Biological; Movement - physiology; Optimization algorithms; Physical Medicine and Rehabilitation; Rotation; Shape; Signal processing; Software; Standard deviation; Studies; Telecommunications and information theory; Tibia - physiology; Validation; Vertebrates: body movement. Posture. Locomotion. Flight. Swimming. Physical exercise. Rest. Sports
• ISSN: 0021-9290; 1873-2380
• DOI: 10.1016/j.jbiomech.2012.08.045

Abstract While measuring knee motion in all six degrees of freedom is important for understanding and treating orthopaedic knee pathologies, traditional motion capture techniques lack the required accuracy. A variety of model-based biplane fluoroscopy techniques have been developed with sub-millimeter accuracy. However, no studies have statistically evaluated the consistency of the accuracy across motions of varying intensity or between degrees of freedom. Therefore, this study evaluated the bias and precision of a contour-based tracking technique by comparing it to a marker-based method (gold standard) during three movements with increasing intensity. Six cadaveric knees with implanted tantalum markers were used to simulate knee extension, walking and drop landings, while motion was recorded by a custom biplane fluoroscopy system. The 3D geometries of the bones were reconstructed from CT scans and anatomical coordinate systems were assigned. The position and orientation of the bone and marker models were determined for an average of 27 frames for each trial and knee joint kinematics were compared. The average bias and precision was 0.01±0.65° for rotations and 0.01±0.59 mm for joint translations. Rotational precision was affected by motion ( p =0.04) and depended on the axis of rotation ( p =0.02). However, the difference in average precision among motions or axes was small (≤0.13°) and not likely of consequence for kinematic measurements. No other differences were found. The contour-based technique demonstrated sub-millimeter and sub-degree accuracy, indicating it is a highly accurate tool for measuring complex three dimensional knee movements of any intensity.