DIAGNOSTIC ACCURACY OF THE MRI IN LUMBAR SPINAL STENOSIS AND LUMBAR DISK HERNIATION.

Abstract

PURPOSE: Toevaluate the diagnostic accuracy of MRI in lumbar spinal stenosis and disk herniation in symptomatic patients by using surgery as the comparison method.

MATERIALS AND METHODS: Sagittal T1-weighted, sagittal and axial T2-weighted MR images were obtained in 32 patients with lumbar spinal stenosis and 18 patients with lumbar disk herniation. All patients underwent surgery. The MR images were evaluated with regarding to intervertebral disk abnormalities, stenosis of the spinal canals, facet joint hypertrophy, ligamentum flavum hypertrophy and nerve root compression. The MR findings were compared with the surgical findings.

RESULTS: Theaccuracy of MRI in diagnosis of disk protrusion, extrusion, and sequestration was 89%, 83% and 89% respectively. Its diagnostic accuracy in detection of herniation location was 86%, 87%, and 89% respectively for central canal, centrolateral location and lateral recess. There was no negative surgical finding in this study. Nerve root compression was 80% accurately diagnosed by MRI. For the 11 cases in which disagreement between MRI and surgical findings, the breakdown was as follows. In 2 patients the MRI and diagnosis was that of disk protrusion but the surgical findings were those of extrusion. In one case, MR diagnosed disk extrusion but the surgery revealed a protruded disk. In one case, there was a sequestrated disk, while MR was interpreted as an extruded disk, and vice versa. About the location of herniation, MR was read as centrolateral location but the surgery showed only lateral location in one case. MR missed 2 cases of centrolateral disk herniation by being interpreted as lateral herniation. Concerning nerve root compression, MR over diagnosed one case of nerve root compression in a surgically proven nerve root deviation. However, MR missed one case of nerve root compression by being read as no compression. The accuracy of MRI in detection of central, centrolateral and lateral recess stenosis was 90%, 89%, and 83% respectively. Its diagnostic accuracy in facet joint hypertrophy, ligamentum flavum hypertrophy and nerve root compression was 83%, 84% and 82% in order. The disagreement between MRI interpretation and surgical findings did occur. Among these 10 patients, MRI showed one patient with centrolateral stenosis but surgery demonstrated only central canal stenosis, and vice versa. In one case, the facet joints were noted as unremarkable during surgery while MRI suggested hypertrophy. In contrary, MRI missed one case of surgically proven facet joint hypertrophy. In one case, MRI was read as ligamentum flavum hypertrophy, while surgery demonstrated normal size. MRI was not able to detect hypertrophy of ligamentum flavum in two cases. MRI over diagnosed one case of nerve root compression in a surgically proven normal nerve root while it missed one case of nerve root compression.

CONCLUSION: MRIis proven to be highly accurate in diagnosis of lumbar disk herniation and lumbar spinal stenosis. However, there is some disagreement on the diagnosis of disk protrusion, extrusion and sequestration; location of herniated disk, and lumbar stenosis; nerve root compression; facet joint hypertrophy and ligamentum flavum hypertrophy.