Verification of measurements of lumbar spinal dimensions in T1- and T2-weighted magnetic resonance imaging sequences.

BACKGROUND CONTEXT Magnetic resonance imaging (MRI) is commonly used to assess patients with lumbar spinal stenosis. No single MRI sequence has been shown to be superior in spinal canal measurements. There are also cost concerns for the increased clinical and research use of MRI. Using only a single sequence may lower the financial burden; however, this requires spinal canal measurements in both T1 and T2 MRI to be reliable. Evidence for this is currently lacking. PURPOSE The aim of this study was to determine the intra- and inter-reader reliability of MRI measurements of the lumbar spine and the reliability of measurements using T1- and T2-weighted MRI films. STUDY DESIGN/SETTING Retrospective study. PATIENT SAMPLE Forty-two randomly selected patients who underwent spinal stenosis surgery. OUTCOME MEASURES Lumbar spinal canal measurements and reliability analysis between T1- and T2-weighted MRI. METHODS Qualitative ratings of MRI features were performed according to previously published criteria by 2 independent readers (JP-YC, HS). Measurements in axial scan included midline anteroposterior (AP) vertebral body diameter, midvertebral body width, midline AP spinal canal diameter, midline AP dural sac diameter, spinal canal width/interpedicular distance, pedicle width (right and left), and lamina angle. Measurements in the sagittal scan included midline AP body diameter, midvertebral body height, and AP spinal canal diameter. Cronbach alpha was used to characterize intra- and inter-reader reliability for qualitative rating data. Similarly, T1 and T2 comparison also was performed in the same manner. RESULTS Good to excellent intra- and interobserver reliability was obtained for all measurements. Reliability analysis of all T1 and T2 measurements was excellent. CONCLUSIONS Either T1 or T2 images can be used for measurements of spinal canal dimensions. These findings are of importance, as not every patient undergoing preoperative MRI assessment will necessarily have both sequences performed and only a single sequence is required for research studies. Our findings are also of relevance in measurement of lumbar canal diameters.