Abstract
Introduction
Adolescent idiopathic scoliosis patients are exposed to ten times more radiation as compared with healthy adolescents, this leads to increased risk of developing malignancies. To minimize the radiation exposure, non-ionizing imaging modalities, such as MRI, can be used, but this is acquired in supine positioning, unlike the standing X-rays that are the gold standard. As far as we know, this is the first study to compare the 2-D morphology of the spine on conventional radiographs with the 3-D morphology on supine MRI and prone CT images. The objective is to evaluate the correlation between the morphology of the spine on conventional radiographs in the upright position and 3-D supine and prone imaging modalities in patients with adolescent idiopathic scoliosis.
Material and Methods
62 adolescent idiopathic scoliosis patients planned for scoliosis surgery and had undergone standard pre-operative work-up (upright radiographs, supine MRI for exclusion of neural axis abnormalities and prone CT for navigation surgery for posterior pedicle screw fixation) were included. In all three positions, Cobb angles, thoracic kyphosis, lumbar lordosis and apical vertebral rotation (2-D X-rays: method of Perdriolle, 3-D scans: semi-automatic software) were determined. In addition, on reconstructed 3-D MR and CT images, the true sagittal and coronal morphology were measured semi-automatically (previously validated technique). Chi-square tests were used for correlation analysis between the positions.
Results
In the thoracic and (thoraco)lumbar curves, Cobb angles correlated significantly between conventional radiographs (68 ± 15° and 44 ± 17°), supine MRI (57 ± 14° and 35 ± 16°) and prone CT (54 ± 15° and 33 ± 15°; r ≥ 0.89; p < 0.001). In the axial plane, the apical vertebral rotation showed a good correlation between the positions (upright, 22 ± 12°; prone, 20 ± 9° and supine, 16 ± 11°; r ≥ 0.56; p < 0.001). The thoracic kyphosis and lumbar lordosis on the conventional upright X-rays did not correlate with the true sagittal morphology on MRI or CT. The relation of the thoracic Cobb angle between the standing X-rays and the lying down images of moderate to severe patients is expressed by the formula: upright radiograph (°) = 10.81 + 1.01*supine MRI(°) and prone CT(°) = −3.02 + 1.00*supine MRI(°).
Conclusion
Although there is slight underestimation of the morphology of the 3-D spinal curvature in the supine position as compared with upright, there is a significant correlation of the deformation in all three dimensions between the different body positions and imaging modalities. Therefore, accurate estimation of the upright morphology of adolescent idiopathic scoliosis is possible, using non-ionizing supine MRI or prone CT scans.