Skip to main content
. Author manuscript; available in PMC: 2014 Feb 15.
Published in final edited form as: Neuroimage. 2012 Nov 21;67:64–76. doi: 10.1016/j.neuroimage.2012.11.014

Table 3.

Test-retest intra-class correlation coefficients (ICC) and coefficients of variation (COV) for DTI parameters in the cervical spinal cord. Slice level variance was factored out in both ICC and COV calculations. Larger ICC value represents small within-subject test-retest variance as compared to between-subject variance, hence higher reproducibility. Smaller COV values represent small within-subject test-retest standard deviation (i.e., square root of variance) as compared to the adjusted mean (Table 2), hence higher reproducibility. Note, ICC did not always correlate with COV; hence these measures provide complementary information about reproducibility. DTI measures with relatively low within-subject variance contribution (ICC≥0.7) or low test-retest variability (COV≤5%) are shown in bold.

FA MD λ λ
ICC WM 0.70 0.50 0.47 0.82
Whole 0.80 0.52 0.60 0.72
GM 0.74 0.52 0.79 0.52
COV (%) WM 3.6 5.9 8.4 8.8
Whole 2.8 6.8 7.5 5.9
GM 6.6 7.5 4.6 10.6