The prediction of lumbar spine geometry: method development and validation

No Thumbnail Available
Issue Date
2005-06
Embargo End Date
Authors
Campbell-Kyureghyan, Naira
Jorgensen, Michael J.
Burr, Deborah
Marras, William S.
Advisor
Citation

Campbell-Kyureghyan, Naira; Jorgensen, Michael J.; Burr, Deborah; Marras, William S. 2005. The prediction of lumbar spine geometry: method development and validation. Clinical Biomechanics, v.20 no.5 pp.455-464

Abstract

Objectives. To develop and validate a new method of predicting the neutral lumbar spine curve from external (non-invasive electrogoniometer) measurements.

Background. Non-invasive techniques for lumbar spine geometry prediction suffer from a lack of a complete geometry description, problems with applicability to field conditions, or both.

Methods.The study consisted of three steps. First, utilizing lateral imaging (MRI and X-ray pictures) of the lumbosacral junction, the torso geometry was described using measures of lumbar lordosis via the Cobb method. Second, the relationship between imaging based measurement of lumbar spinal curvature and externally measured torso flexion angle in the sagittal plane using a goniometer was determined. Finally, method validation was performed with an independent set of nine subjects. The predicted lumbar spine curve was determined and the prediction errors were analyzed against the measured curves from digitized lateral X-ray images of the lumbosacral junction.

Results.The shape of the lumbar curve was described as function of three externally measured parameters. The lumbar spine Cobb angle, segmental centroid positions (S1−T12), and segmental orientations were predicted from the external lumbar motion monitor measurements, with average precisions of 5.8°, 4.4 mm, and 3.9°, respectively.

Conclusions.The position and orientation of each segment (vertebrae and disc), along with the lumbar spine lordosis, can be predicted in the neutral posture using data from back angular measurements.

Table of Content
Description
Click on the DOI link to access the article (may not be free).
publication.page.dc.relation.uri
DOI