Three-dimensional movements of the lumbar spine facet joints and segmental movements: in vivo examinations of normal subjects with a new non-invasive method

Abstract

Introduction: Examination with CT and image registration is a new technique that we have previously used to assess 3D segmental motions in the lumbar spine in a phantom. Current multi-slice computed tomography (CT) offers highly accurate spatial volume resolution without significant distortion and modern CT scanners makes it possible to reduce the radiation dose to the patients. Our aim was to assess segmental movement in the lumbar spine with the aforementioned method in healthy subjects and also to determine rotation accuracy on phantom vertebrae.

Material and method: The subjects were examined in flexion-extension using low dose CT. Eleven healthy, asymptomatic subjects participated in the current study. The subjects were placed on a custom made jig which could provoke the lumbar spine into flexion or extension. CT examination in flexion and extension was performed. The image analysis was performed using a 3D volume fusion tool, registering one of the vertebrae, and then measuring Euler angles and distances in the registered volumes.

Results: The mean 3D facet joint translation at L4-L5 was in the right facet joint 6.1 mm (3.1-8.3), left facet joint 6.9 mm (4.9-9.9), at L5-S1: right facet joint 4.5 mm (1.4-6.9), and for the left facet joint 4.8 mm (2.0-7.7). In subjects the mean angles at the L4-L5 level were: in the sagittal plane 14.3°, coronal plane 0.9° (-0.6 to 2.8), and in the transverse plane 0.6° (-0.4 to 1.5), in the L5-S1 level the rotation was in sagittal plane 10.2° (2.4-16.1), coronal plane 0° (-1.2 to 1.2), and in the transverse plane 0.2° (-0.7 to 0.3). Repeated analysis for 3D facet joint movement was on average 5 mm with a standard error of mean of 0.6 mm and repeatability of 1.8 mm (CI 95%). For segmental rotation in the sagittal plane the mean rotation was 11.5° and standard error of mean 1°. The repeatability for rotation was 2.8° (CI 95%). The accuracy for rotation in the phantom was in the sagittal plane 0.7°, coronal plane 1°, and 0.7 in the transverse plane.

Conclusion: This method to assess movement in the lumbar spine is a truly 3D method with a high precision giving both visual and numerical output. We believe that this method for measuring spine movement is useful both in research and in clinical settings.

Figures

Fig. 1

A subject on the jig in extension provocation

Fig. 2

A subject on the jig in flexion provocation

Fig. 3

2D slice image of the L5 vertebra after registration. The window is split so that the bottom half is the reference volume, and the top half is the registered volume

Fig. 4

3D isosurface representation of L5 after registration. The reference and registered volume are represented by different colours. Note the “zebra-like” interference pattern between the colours, indicating a good match. This pattern arises when the overlap is better than the smallest image element (the voxel). These images can be viewed from an arbitrary angle

Fig. 5

Close-up of the right L4 facet joint in a 3D volume. The landmarks are all placed on the L4 side of the facet joint; one in the most cranial point, one in the most caudal point, one in the most anterior point and one posterior point in the periphery of the facet

Fig. 6

Facet joint movement in L4–L5 represented in 3D. Note the Zebra pattern in L5 that indicate a almost perfect match

Fig. 7

On-screen representation of a L4–L5 flexion between measured to 9.8° in the sagittal plane. 3D isosurface display. The volume colours are reduced so they are more transparent which makes it easy to see the movement in L4. Note that L5 looks like a single vertebra due to an almost perfect match (fused)