Quantitative magnetic resonance imaging assessment of lateral atlantoaxial joint meniscoid composition: a validation study

Abstract

2 AbstractPurpose: Lateral atlantoaxial (LAA) joint meniscoid composition may have clinical significance in patients following neck trauma. However, the existing method of radiologically assessing meniscoid composition has an inherent element of subjectivity, which could contribute to measurement variability. The present study sought to investigate the accuracy of 2 point Dixon fat/water separation MRI as a quantitative assessment of LAA joint meniscoid composition. Methods: Sixteen LAA joint meniscoids were excised from four cadavers (mean [SD] age 79.5 [3.7] years; one female) following cervical spine MRI (2-point Dixon, T1-weighted VIBE and T2-weighted SPACE sequences). Composition of LAA joint meniscoids was undertaken by i) histological examination by light microscopy, ii) calculation of fat fraction by Dixon MRI (both In-phase/Opposedphase and Fat/Water methods), and iii) the existing method of considering VIBE and SPACE signal intensities. Analysis was performed using the kappa statistic with linear weighting. Results: Microscopy revealed three, five and eight meniscoids to be composed of adipose, fibroadipose and fibrous tissues, respectively. Dixon sequence MRI classified 11 of these meniscoids correctly, with 'substantial' level of agreement (In-phase/Opp-phase kappa statistic = 0.78 [95% CI 0.38, 1.17]; Fat/Water kappa statistic = 0.72 [95% CI 0.32, 1.11]). Level of agreement between microscopy and the VIBE and SPACE method was 'poor' (kappa statistic = 0.02 [95% CI -0.34, 0.38]).Conclusions: Findings suggest that Dixon fat/water separation MRI may have superior utility in the assessment of LAA joint meniscoid composition than the existing method of considering VIBE and SPACE signal intensities.