A retrospective cohort study.Objective.
Scoliosis and hydromyelia have frequently been observed in patients with spinal dysraphism. We investigated the applicability of curved planar reformation (CPR) for evaluating hydromyelia in patients with scoliosis associated with spinal dysraphism.Summary of Background Data.
It is quite difficult to evaluate scoliosis and hydromyelia in patients with spinal dysraphism.Methods.
We identified 11 patients with scoliosis and a Cobb angle of >20° among 107 spinal dysraphism patients. In addition to routine T1- and T2-weighted axial and sagittal MRI, we obtained three-dimensional constructive interference in steady-state magnetic resonance imaging (MRI) as sagittal cross-section volume images. The spinal cord and hydromyelia were rendered into a single-plane, two-dimensional image using the straightened CPR technique. In cases of scoliosis with hydromyelia, the sagittal length and maximal diameters of hydromyelia and the spinal cord were measured by three examiners. Measurement errors among examiners were evaluated using standard deviation (SD) and coefficient of variation (CV).Results.
Each series of image sets provided a straightened CPR image that clearly delineated the entire length of the linearized spinal cord in a single plane. The straightened CPR image also demonstrated the accurate lengths and estimated volume of hydromyelia. Given that three of the 11 patients underwent serial MRI during this period, we were able to accurately compare volume changes. In the sagittal length of the hydromyelia, SD and CV were smaller with the straightened CPR technique than with the conventional T2 WI sagittal section in all cases, showing a statistically significant difference between both techniques (SD: P = 0.014, CV: P = 0.013). Even in the measurements, the difference in CV between both techniques was close to statistical significance.Conclusion.
The straightened CPR technique is useful for accurately identifying volume changes in hydromyelia, even in patients with severe scoliosis.Conclusion.
Level of Evidence: 1