Abstract
Sheep are established models for intervertebral disc degeneration (IVDD) translational research. Objectives of this retrospective, secondary analysis, observational study were to evaluate reproducibility of two magnetic resonance imaging (MRI) measures of IVDD in sheep using open-source image analysis software. Sagittal planar, T2-weighted, lumbar MRI scans from a previous, prospective study were evaluated (4 sheep @ 3 treatment periods/sheep @ 7 lumbar discs /treatment period = 84 discs). A standardized protocol was developed for measuring MRI index (nucleus pulposus signal intensity × nucleus pulposus sagittal area), and nucleus pulposus signal intensity standard deviation (NPSISD). Scans were randomized and four observers with varying experience levels applied the protocol to independently perform triplicate measures. Mean coefficient of variation values for both the MRI index and NPSISD were within an acceptable range (< 10). Mean values for the three readers with at least 2 months of image analysis experience did not differ (P > .05).