Quantitative measurement of intervertebral disc signal using MRI

To investigate the spinal cord as an alternative intra-body reference to cerebrospinal fluid (CSF) in evaluating thoracic disc signal intensity.

T2-weighted magnetic resonance imaging (MRI) images of T6-T12 were obtained using 1.5 T machines for a population-based sample of 523 men aged 35-70 years. Quantitative data on the signal intensities were acquired using an image analysis program (SpEx). A random sample of 30 subjects and intraclass correlation coefficients (ICC) were used to examine the repeatability of the spinal cord measurements. The validity of using the spinal cord as a reference was examined by correlating cord and CSF samples. Finally, thoracic disc signal was validated by correlating it with age without adjustment and adjusting for either cord or CSF. Pearson's r was used for correlational analyses.

The repeatability of the spinal cord signal measurements was extremely high (>or=0.99). The correlations between the signals of spinal cord and CSF by level were all above 0.9. The spinal cord-adjusted disc signal and age correlated similarly with CSF-adjusted disc signal and age (r=-0.30 to -0.40 versus r=-0.26 to -0.36).

Adjacent spinal cord is a good alternative reference to the current reference standard, CSF, for quantitative measurements of disc signal intensity. Clearly fewer levels were excluded when using spinal cord as compared to CSF due to missing reference samples.