Current assessment of spinal degenerative disease with magnetic resonance imaging

Radiography (plain roentgenography, myelography, computed tomography (CT), computed tomographic myelography) has been used to identify morphologic changes involving the various components of the diskovertebral unit. Added to this armamentarium of imaging techniques is magnetic resonance (MR) imaging, with its superior ability to define anatomy, its improved contrast sensitivity, and its potential to provide unique biochemical and physiologic information. The authors review the current use of MR imaging in defining degenerative changes in the spine including the various patterns of herniation, annular tears, canal stenosis, and the use of gadolinium-diethylenetriamine-pentaacetic acid for previously unoperated and operated patients. Prospective studies have compared surface-coil MR imaging, CT, and myelography in the evaluation of disk herniation and stenosis and found an 82.6% accuracy between MR imaging and surgical findings for the type and location of the disease. Recent experience with precontrast and postcontrast MR imaging in the postoperative lumbar spine indicated that it was 96% accurate in differentiating scar from disk in 44 patients at 50 reoperated levels. Three-dimensional imaging is, more and more, becoming an integral part of routine MR imaging. The theoretical and practical advantages of three-dimensional imaging are several and include a theoretical increase in the signal-to-noise ratio over two-dimensional imaging (by the square root of the number of partitions selected), the ability to obtain thin contiguous slices from the volume without the problem of cross-talk found in two-dimensional imaging, more accurate slice thickness than that achieved in two-dimensional imaging, and a reduction in susceptibility artifacts. Different three-dimensional techniques are capable of providing either high or low signal intensity cerebrospinal fluid (CSF), with excellent suppression of CSF pulsation artifacts. Certain sequences provide a high enough signal intensity that a computer algorithm may be used to display the CSF in a rotating three-dimensional manner, similar to a myelogram. This three-dimensional myelographic image has the potential of providing the clinician with a global assessment of the CSF spaces, an advantage previously lacking with other imaging techniques.