Agreement and Repeatability of Linear Vertebral Body and Canal Measurements Using Computed Tomography (CT) and Low Field Magnetic Resonance Imaging (MRI)

Abstract

To evaluate agreement and repeatability of vertebral column measurements using computed tomography (CT) and magnetic resonance imaging (MRI). Retrospective observational study. ANIMALS; Dogs (n=18) with disc associated wobbler syndrome; Dog cadavers (n=3). Five measurements of the 5th cervical vertebra were performed: vertebral body length (VBL), vertebral canal height (VCH), vertebral body height (VBH), vertebral canal width (VCW), and vertebral body width (VBW). Measurements were performed independently twice by 2 observers. Bland-Altman plots were created to evaluate agreement. Cadaveric vertebrae with soft tissue removed had the same variables and actual dimensions measured. The largest discrepancy between CT and MRI measurement was for VBL (mean difference+/-SD=1.262 mm+/-1.245; P<.001), with the difference for all the other variables being acceptable. The 1st measurement was significantly higher than the 2nd only for VBL using CT (mean difference=0.476 mm+/-1.120; P=.009), with all other variables having acceptable differences. Mean difference for all measurements between 2 observers was small, except for VBL using CT (mean difference=0.762 mm+/-1.042; P<.001). Only the difference for VBL between CT and cadaver specimens was statistically significant. Our results suggest high repeatability and good agreement for most vertebral measurements of interest. VBL measurement using CT was considered problematic. Provided limitations are understood, linear measurements of vertebral dimensions from CT and MRI images can be used clinically.