Dixon Imaging of Bone Marrow Lesions in Osteoarthritis of The Knee

Abstract

Abstract: Aim: This study compared between SPAIR and 3D gradient echo quantitative Dixon imaging for measuring subchondral BMLs in KOA. Moreover, it aims to compare the two sequences in measuring BMLs volume at baseline and changes over time to determine if the Dixon sequence may be a feasible alternative to the SPAIR sequence in KOA. Method: The sample (n=17) included (9 females and 8 males) with knee OA. Obtained MRI at baseline and follow-up (the median time was 24-36 weeks). SPAIR and 3D GE Dixon sequences were obtained using a (3 T scanner). For measurements of BMLs in KOA. A total of BMLs volume for 68 images were identified as high MRI intensity and MS by using the "Osirix MD" software (v.12.0.3) to draw ROI around BMLs for each image. Some statistical analysis were calculated such as The average mean, Paired t-test (parametric test) and Pearson correlation coefficient for comparison purposes. Moreover, the 68 sagittal SPAIR and 3D GE Dixon images were assessed by one observer therefore, the ICC was undertaken to assess repeatability the first measurement from second measurement from the same images. Result: the mean volumes of BMLs were similar with no significant difference for SPAIR and Dixon images at baseline and change over time. The baseline correlation was strong (0.73) whereas the correlation over time was weak (0.27). However, and the intra-observer ICC were higher at baseline and change over time. Conclusion: Dixon imaging can be a viable alternative to SPAIR for baseline, because it provided relatively similar results and strong correlation with SPAIR. However, Dixon imaging cannot replace SPAIR for measuring changes of BMLs over time because the results showed weak correlation between them