Comparison of magnetic resonance imaging and computed tomogram-myelography for evaluation of cross sections of cervical spinal morphology.

Abstract

Comparison of magnetic resonance imaging (MRI) and computed tomography-myelography (CTM) for cervical intracanalar dimensions. To compare the capability and reproducibility of MRI and CTM in measuring the cross-sectional morphology of intracanalar lesions of the cervical spine. The relative advantages and disadvantages of MRI and CTM in measuring cervical intracanalar dimensions are poorly understood. MRI and CTM were used to measure cervical disc levels in 45 subjects with various cervical spinal diseases. Measurements included dural area, dural anteroposterior (A-P) diameter, dural right-left (RL) diameter, cord area, cord anteroposterior (A-P) diameter, cord right-left (RL) diameter and cerebrospinal fluid (CSF) space (anterior and posterior). Each section was graded by 2 orthopedic surgeons for degree of stenosis (Grades, 0-3), and the intra- and interobserver reproducibility of these measurements (intraclass correlation coefficients: ICC) was assessed. In both CTM and MRI, intra- and interobserver reproducibility (ICC) ranged from 0.702 to 0.989, suggesting that both imaging methods are reproducible. Importantly, CTM measurements of dural area, dural A-P diameter, dural RL diameter, and CSF space (anterior and posterior) were slightly, but significantly (P < 0.001), larger than MRI measurements. In contrast, MRI measurements of cord area, cord A-P diameter, and cord RL diameter were slightly, but significantly (P < 0.001), larger than CTM measurements. Degree of stenosis was significantly more severe in MRI than in CTM. Both CTM and MRI provided reproducible measurements of cervical intracanalar dimensions. Measurements of dura were slightly larger in CTM, whereas measurements of spinal cord were slightly larger in MRI, making stenosis more severe in MRI than in CTM. The clinical relevance of these slight differences requires further examination.