Assessment of magnetic resonance imaging in the diagnosis of lumbar spine foraminal stenosis--a surgeon's perspective.

STUDY DESIGN A 2-part cadaveric study. OBJECTIVES Part 1: To assess the reliability of a semiquantitative classification system for grading lumbar spine foraminal stenosis and the variability in magnetic resonance imaging (MRI) examinations carried out in different institutions. Part 2: to assess the difference between the foraminal measurements carried out on MRI images and on cadaveric specimens. SUMMARY OF BACKGROUND DATA There are limited data to suggest that MRI examinations are sensitive or specific for the diagnosis of lumbar spine foraminal stenosis. The effect of the variability in techniques, used by different imaging centers, is not well understood. METHODS Three fresh, frozen human lumbar spines were examined with 3 different clinical MRI systems. Eight spine surgeons used the MRI images to grade the foramina, on the basis of a previously suggested semiquantitative classification. In addition, the dimensions of all foramina were measured using software available on each of the MRI systems. The dimensions of the specimens' foramina were then measured in situ, graded, and compared with the MRI measurements. RESULTS AND CONCLUSIONS There was poor intraobserver agreement using the semiquantitative grading system. The parameters associated with the grade of stenosis assigned to the foramen were as follows: (1) the observer doing the grading, (2) the place it was imaged, and (3) the location of the foramen. There was poor correlation between measurements of the foramina carried out on MRI and the specimens.