Abstract
Introduction
Increasing studies on the muscles volume in the lumbar spine of low back pain patients is available in the literature. So far, these studies were focused only on the volume of the muscle, but the quality of muscle is more important, and the quality of muscle can be evaluated with the fat fraction ratio. To measure fat fraction of muscle, currently, 2, 3 points echo Dixon techniques were used, but its accuracy is a problem. Therefore, we use the first 6 points echo Dixon and to compare accuracy of 2 echo Dixon, 3 point Dixon imaging with that from MR spectroscopy as reference.
Patients and Methods
A total of 69 patients (39 women and 22 men; 54.3 ± 19.1) underwent MRI at a 1.5 T scanner. In addition, using T2-corrected Dixon sequence with 2, 3, and 6 echoes were obtained at the L4–L5 levels for fat-signal fraction. Fat-signal fraction from MR spectroscopy was also obtained at lumbar muscles of L4–L5 levels. Fat-signal fractions were measured directly by drawing region of interest at the automatically obtained fat-signal fraction mapping images from the three sequences. ROIs were drawn at the three consecutive slices of mapping images at the same location of the spectroscopic voxel by two musculoskeletal radiologists in consensus and the average values were obtained. The Student t test and Bland–Altman plots were used to quantify agreement between the values obtained from mapping images and those from spectroscopy. The p values < 0.05 were considered to be statistically significant.
Results
A total of 120 spectroscopic measurements were performed bilaterally. Mean spectroscopic fat-signal fraction percentage was 14.0 ± 11.8 (range, 2.9–63.6). Correlation between spectroscopic and all imaging-based fat-signal fractions was statistically significant (R2 = 0.92 [2 echo], 0.91 [3 echo], and 0.96 [6 echo], all p < 0.001). Fat-signal fractions obtained from 6-echo T2-corrected Dixon sequence best correlated among all imaging-based fat-signal fractions with statistical significance (p < 0.001).
Conclusion
T2-corrected 6-echo Dixon sequence best correlates with spectroscopic fat-signal fractions as compared with 2 echo and 3 echo sequences, thus being expected to be used as an accurate quantification tool for measurement of muscle fat quantification in lumbar spine MR imaging.