Assessment of The Accuracy of The MR Abdominal Adipose Tissue Volumetry using 3D Gradient Dual Echo 2-Point DIXON Technique using CT as Reference

Abstract

In this study, in order to determine the validity and accuracy of MR imaging of 3D gradient dual echo 2-point DIXON technique for measuring abdominal adipose tissue volume and distribution, the measurements obtained by CT were set as a reference for comparison and their correlations were evaluated. CT and MRI scans were performed on each subject (17 healthy male volunteers who were fully informed about this study) to measure abdominal adipose tissue volume. Two skilled investigators individually observed the images acquired by CT and MRI in an independent environment, and directly separated the total volume using region-based thresholding segmentation method, and based on this, the total adipose tissue volume, subcutaneous adipose tissue volume and visceral adipose tissue volume were respectively measured. The correlation of the adipose tissue volume measurements with respect to the observer was examined using the Spearman test and the interobserver agreement was evaluated using the intra-class correlation test. The correlation of the adipose tissue volume measurements by CT and MRI imaging methods was examined by simple regression analysis. In addition, using the Bland-Altman plot, the degree of agreement between the two imaging methods was evaluated. All of the statistical analysis results showed highly statistically significant correlation (p<0.05) respectively from the results of each adipose tissue volume measurements. In conclusion, MR abdominal adipose volumetry using the technique of 3D gradient dual echo 2-point DIXON showed a very high level of concordance even when compared with the adipose tissue measuring method using CT as reference.

Advanced Molecular Imaging (MRI/MRS/1H NMR) for Metabolic Information in Young Adults with Health Risk Obesity

“…A complete ectopic fat estimation is done by the total breakdown of net MR signals into fat and water separately in MRI. Moreover, MR results show a very high agreement concordance with CT in abdominal fat estimation [24]. The chemical shift-based MRS shows several metabolites from a single test.…”

“…This combination of manual and automated segmentation methods quantified the different abdominal fat distributions. Abdominal subcutaneous fat percentage (Sub fat %) was also calculated by subtracting Vis fat % from total abdominal fat percentage (Abd fat %) [24]. (G-fat %), soleus (S-fat %), and tibialis anterior (TA-fat %).…”