Ergen Fatma Bilge, Gulal Gulsah, Yildiz Adalet Elcin, Celik Azim, Karakaya Jale, Aydingoz Ustun
From the *Department of Radiology, Hacettepe University School of Medicine, Ankara, Turkey; † Department of Radiology, Göztepe Teaching and Research Hospital, Istanbul, Turkey; ‡GE Healthcare, Istanbul, Turkey; and §Department of Biostatistics, Hacettepe University School of Medicine, Ankara, Turkey.
J Comput Assist Tomogr. 2014 Mar-Apr;38(2):320-4. doi: 10.1097/RCT.0b013e3182aa4d9d.
The aim of this study was to use the T2*-iterative decomposition of water and fat with echo asymmetry and least squares estimation quantification (IDEAL) technique to estimate vertebral fat fraction (FF) and compare it with bone mineralization levels in females.
Forty-five (mean age, 49.5 years) consecutive women who underwent magnetic resonance imaging and spinal dual-energy x-ray absorptiometry bone mineral density (BMD) examination constituted the study population. Depending on t scores derived from dual-energy x-ray absorptiometry, the vertebrae were classified into 3 groups (1, healthy; 2, osteopenia; and 3, osteoporosis). The Spearman ρ test was used to investigate the presence of correlation between FF and BMD. Analysis of covariance was performed to compare the differences among the groups. The FF cutoff value for the prediction of osteoporosis/osteopenia was evaluated with the receiver operating characteristic curve analysis.
We found a moderate negative correlation between BMD (grams per square centimeter) and FF (r = - 0.42), and it was statistically significant (P < 0.001). The FF mean of age-corrected group 1 (45.17%; SD, 1.3) was lower than that of groups 2 (51.77%; SD, 0.69) and 3 (50.82%; SD, 1.5), and the difference was statistically significant (P < 0.001, P = 0.021). The area under the receiver operating characteristic curve for FF was 0.80 (95% confidence interval, 0.72-0.86). The optimal cutoff point was obtained as 39%, and for this cutoff point, the sensitivity and the specificity were 93% and 60.3%, respectively.
The T2*-IDEAL technique can be used as an alternative technique in estimation of FF, and it is possible to detect reduced bone mineralization of the vertebrae by estimation of FF value with this technique.
本研究旨在使用具有回波不对称性和最小二乘估计量化（IDEAL）技术的水和脂肪的T2*迭代分解来估计椎体脂肪分数（FF），并将其与女性的骨矿化水平进行比较。
连续45名（平均年龄49.5岁）接受磁共振成像和脊柱双能X线吸收法骨密度（BMD）检查的女性构成研究人群。根据双能X线吸收法得出的t值，将椎体分为3组（1组，健康；2组，骨量减少；3组，骨质疏松）。使用Spearman ρ检验来研究FF与BMD之间的相关性。进行协方差分析以比较各组之间的差异。通过受试者工作特征曲线分析评估预测骨质疏松症/骨量减少的FF临界值。
我们发现BMD（每平方厘米克数）与FF之间存在中度负相关（r = -0.42），且具有统计学意义（P < 0.001）。年龄校正后的1组FF平均值（45.17%；标准差，1.3）低于2组（51.77%；标准差，0.69）和3组（50.82%；标准差，1.5），差异具有统计学意义（P < 0.001，P = 0.021）。FF的受试者工作特征曲线下面积为0.80（95%置信区间，0.72 - 0.86）。最佳临界点为39%，对于该临界点，敏感性和特异性分别为93%和60.3%。
T2* - IDEAL技术可作为估计FF的替代技术，通过该技术估计FF值有可能检测到椎体骨矿化减少。
