Effect of intervertebral changes of the spinal trabecular and cortical mineral content on the precision requirements in longitudinal single and dual energy computed tomography examinations.

Statistical considerations related to the measuring and analyzing procedures in computed tomography (CT) scans indicate that determination of a 2.5% change in spinal mineral content (SMC) which is assumed necessary in clinical practice, requires a 0.9% combined precision in setting the location and direction of the scan plane through the vertebral body, and defining the region of interest. In single energy examinations the short-term precision of CT scanners was found to be sufficiently high to achieve this goal, but the variability due to subjective finding of scan direction requires the automation of slice selection and additional averaging of data from more than one scan slice. The precision of calcium images was much lower, and one conclusion of this study is that dual energy technology is not suitable for longitudinal examinations, where small changes in SMC have to be determined. The SMC of the vertebrae T12, L1, L2, and L3 as measured in 41 subjects was compared and a significant decrease in trabecular bone in single energy examinations and increase in cortical bone was found from T12 to L3. This suggests that the comparison of SMC data should be done over time for each vertebra separately, in order to preserve diagnostic information.