Thomas R Ward1, Mafruha M Hussain2, Mark Pickering2, Diana Perriman1, Al Burns1, Jennie Scarvell1,3, Paul N Smith1,4. 1. Trauma and Orthopaedic Research Unit, Canberra Hospital, ACT, Australia. 2. University of New South Wales at The Australian Defence Force Academy, Australia. 3. The University of Canberra, Australia. 4. The Australian National University, Canberra, Australia.
Abstract
INTRODUCTION: A kinematic measurement method combining dynamic motion and imaging, which captures the behaviour of the hip at terminal motion, may offer improved diagnostic accuracy and enhance our understanding of the mechanics of femoroacetabular impingement (FAI). METHODS: 3 embalmed cadaveric hip/pelvis specimens with implanted Roentgen Stereophotogrammetric Analysis (RSA) beads were mounted on a custom rig and imaged with a fluoroscope in four poses to simulate a clinical impingement examination: in hip extension and in three positions: near impingement, early impingement and late impingement while simulating a flexion/adduction/internal rotation manoeuvre. Hip joint kinematics were measured using 2 methods and compared: RSA (gold standard) and a custom 3-dimensional to 2-dimensional (3D-2D) image registration method which matches 3D models developed from CT to 2D fluoroscopic images. RESULTS: Using RSA as the gold standard, bias and precision of hip joint rotations measured using 3D-2D registration demonstrated maximums of 1.64° and 3.96°, respectively. However, if the single outlier was removed, bias and precision were 0.55° and 1.38°. Bias and precision of translations had maximums of 0.51 mm and 0.77 mm, respectively. CONCLUSIONS: This 3D to 2D registration method may offer a clinically useful solution for dynamic assessment of hip impingement. If 5-mm translation and 10° of rotation represent a clinically significant difference in hip kinematics, the method's accuracy of approximately 1 mm displacement and 1° rotation should enable detection of significant clinical differences.
INTRODUCTION: A kinematic measurement method combining dynamic motion and imaging, which captures the behaviour of the hip at terminal motion, may offer improved diagnostic accuracy and enhance our understanding of the mechanics of femoroacetabular impingement (FAI). METHODS: 3 embalmed cadaveric hip/pelvis specimens with implanted Roentgen Stereophotogrammetric Analysis (RSA) beads were mounted on a custom rig and imaged with a fluoroscope in four poses to simulate a clinical impingement examination: in hip extension and in three positions: near impingement, early impingement and late impingement while simulating a flexion/adduction/internal rotation manoeuvre. Hip joint kinematics were measured using 2 methods and compared: RSA (gold standard) and a custom 3-dimensional to 2-dimensional (3D-2D) image registration method which matches 3D models developed from CT to 2D fluoroscopic images. RESULTS: Using RSA as the gold standard, bias and precision of hip joint rotations measured using 3D-2D registration demonstrated maximums of 1.64° and 3.96°, respectively. However, if the single outlier was removed, bias and precision were 0.55° and 1.38°. Bias and precision of translations had maximums of 0.51 mm and 0.77 mm, respectively. CONCLUSIONS: This 3D to 2D registration method may offer a clinically useful solution for dynamic assessment of hip impingement. If 5-mm translation and 10° of rotation represent a clinically significant difference in hip kinematics, the method's accuracy of approximately 1 mm displacement and 1° rotation should enable detection of significant clinical differences.