Magnetic resonance imaging of the lumbar spine with CT correlation

The results of magnetic resonance (MR) imaging and computed tomography (CT) in 18 patients with known degenerative disk disease of the lumbar spine were compared. In 60 intervertebral disk levels studied, there were 17 disks with degeneration and disk bulge, and 15 herniated disks. Final diagnoses were based on several factors, with surgical confirmation in five patients. There was good correlation between the two methods at 51 of the 60 levels studied. However, there were major discrepancies in interpretation at nine intervertebral disk levels. These included three false-positive MR imaging interpretations of a herniated disk and one false-negative herniated disk on MR imaging. MR imaging detected one case of disk herniation that was missed prospectively on CT. There were also four presumed degenerated disks seen on MR scans that appeared normal on CT. The conus medullaris was imaged in 16 of 18 patients. The sagittal view proved best for demonstrating both disk abnormality and the conus medullaris. The transaxial view was sometimes helpful in localizing a disk herniation, but partial-volume averaging in the 7-mm slice thickness limited its usefulness. There were five disk herniations that could not be accurately localized on the MR scan. MR imaging proved more sensitive than CT in detecting early disk disease, which appeared as decreased signal intensity within the disk. In three postoperative cases, MR imaging was better able to distinguish between recurrent disk herniation and postoperative scar formation. CT, on the other hand, was more specific in distinguishing herniated disk from disk bulge and proved far superior to MR imaging in localizing disk herniation.(ABSTRACT TRUNCATED AT 250 WORDS)