Intra- and inter-rater reliability for the measurement of the cross-sectional area of ankle tendons assessed by magnetic resonance imaging

Background The cross-sectional area (CSA) records make an essential measurement for determining the mechanical properties of tendons, such as stress and strength. However, there is no consensus regarding the best method to record the CSA from different tendons. Purpose To determine intra- and inter-rater reliability for CSA measures from magnetic resonance imaging (MRI) of the following tendons: tibialis anterior; tibialis posterior; fibularis longus and brevis; and Achilles. Material and Methods We designed an observational study with repeated measures taken from a convenience sample of 20 participants diagnosed with acute or chronic ankle sprain. Two independent raters took three separate records from the CSA of ankle tendon images of each MRI slice. The intra-class correlation coefficient (ICC) and 95% limits of agreement (LoA) defined the quality (associations) and magnitude (differences), respectively, of intra- and inter-rater reliability on the measures plotted by the Bland-Altman method. Results Data showed very high intra- and inter-rater correlations for measures taken from all tendons analyzed (ICC 0.952-0.999). It also revealed an excellent agreement between raters (0.12%-2.3%), with bias no higher than 2 mm(2) and LoA in the range of 4.4-7.9 mm(2). The differences between repeated measures recorded from the thinnest tendons (fibularis longus and brevis) revealed the lowest bias and narrowest 95% LoA. Conclusion Reliability for the CSA of ankle tendons measured from MRI taken by independent rates was very high, with the smallest differences between raters observed when the thinnest tendon was analyzed.