Quantification of intervertebral efforts for assisted scoliotic spine surgeries : accuracy of preoperative manual identification of vertebral landmarks

Details

Supervisors

Cartiaux, Olivier ; Abedrabbo, Gabriel

Faculty

Faculté des sciences de la motricité

Degree label

Master [120] en sciences de la motricité, orientation générale

Abstract

Surgical treatment of scoliosis can cause decrease in spine motion and lead to early degenerative changes in adjacent segments depending the extension of the fusion. Quantification of intervertébral efforts may assist surgeon in deciding fusion extent. The 3D spine geometry is obtained from biplane x-rays and is used for the kinematics acquisition and intervertébral efforts quantification. The aim of this study was to investigate the extent and the sources of error in the spine reconstruction using conventional x-rays. An in vitro experimental comparison was performed using a synthetic spine placed in a healthy and a scoliotic configuration. A reference frame was used to scale X-ray images in order to correct radiographie magnification. The reconstruction based on X-rays of the spine was compared to the reference obtained with a 3D digitizer. Results showed mean errors in X- and Y-directions of 2,8mm for the healthy spine and of 0,97mm for the scoliotic spine. The major finding was the errors in the Z-direction increasing with the vertebrae level, with mean errors of 2,1mm and 8mm for respectively dise centers and spinous processes of the healthy spine and 4,8mm and 6,4mm for the scoliotic one. This errors seem to be largely influenced by the coronal distance between the reference frame and the anatomical landmarks. A true reconstruction of the spine using conventional X-ray acquisition is hard to achieve. This study brings out some of the potential sources of error and their impact in the accuracy of the spine geometry model.