Three-dimensional dynamic in vivo motion of the cervical spine: assessment of measurement accuracy and preliminary findings

BACKGROUND CONTEXT: Previous research has quantified cervical spine motion with conventional measurement techniques (eg, cadaveric studies, motion capture systems, and fluoroscopy), but these techniques were not designed to accurately measure three-dimensional (3D) dynamic cervical spine motion under in vivo conditions. PURPOSE: The purposes of this study were to characterize the accuracy of model-based tracking for measuring 3D dynamic cervical spine kinematics and to demonstrate its in vivo application. STUDY DESIGN: Through accuracy assessment and application of technique, in vivo cervical spine motion was measured. METHODS: The accuracy of model-based tracking for measuring cervical spine motion was determined in an in vitro experiment. Tantalum beads were implanted into the vertebrae of an ovine specimen, and biplane X-ray images were acquired as the specimen's neck was manually moved through neck extension and axial neck rotation. The 3D position and orientation of each cervical vertebra were determined from the biplane X-ray images using model-based tracking. For comparison, the position and orientation of each vertebra were also determined by tracking the position of the implanted beads with dynamic radiostereometric analysis. To demonstrate in vivo application of this technique, biplane X-ray images were acquired as a human subject performed two motion tasks: neck extension and axial neck rotation. The positions and orientations of each cervical vertebra were determined with model-based tracking. Cervical spine motion was reported with standard kinematic descriptions of translation and rotation. RESULTS: The in vitro validation demonstrated that model-based tracking is accurate to within +/- 0.6 mm and +/- 0.6 degrees for measuring cervical spine motion. For the in vivo application, there were significant rotations about all three anatomical axes for both the neck extension and axial neck rotation motion tasks. CONCLUSIONS: Model-based tracking is an accurate technique for measuring in vivo, 3D, dynamic cervical spine motion. Preliminary data acquired using this technique are in agreement with previous studies. It is anticipated that this experimental approach will enhance our understanding of cervical spine motion under normal and pathologic conditions. (C) 2010 Elsevier Inc. All rights reserved.