Vertebral Bone Marrow Fat Quantification and its Relationship with Bone Mineral Density: Using Multi-Echo MRS and Multi-Echo Dxion
Na Chai1, Panli Zuo2, Stephan Kannengiesser3, Andre De Oliveira3, Shun Qi1, and Hong Yin1

1Department of Radiology, Xijing Hospital, Xi'an, China, People's Republic of, 2Siemens Healthcare, MR Collaborations NE Asia, Beijing, China, People's Republic of, 3Application Predevelopment, Siemens Healthcare, Erlangen, Germany


Using the multi-echo 1H-MRS and multi-echo Dixon VIBE, we measured the proton density fat fraction (PDFF) using MR imaging in the bone marrow of L2-L4 vertebra, and compared with the bone mineral density (BMD) measured using computed tomography (CT). The resutls showed a significant correlation between PDFF measured using the two methods, and also PDFF with BMD.


Bone mineral density (BMD) measured using computed tomography (CT) has been used to determine fracture risk and potential for osteoporosis in clinical practice. In previous studies, vertebral bone marrow fat fraction showed a reliable correlation with BMD using single-voxel proton magnetic resonance spectroscopy (1H-MRS) with single-echo. However, the difference in T2-relaxation between water and fat induces a bias in the quantification of fat fraction (FF) at different TE. In order to get the proton density fat fraction (PDFF) without being influenced by other factors, in this study we used multi-echo 1H-MRS and multi-echo Dixon to measure the PDFF in L2-L4 spines and compared it with corresponding BMD measured by CT.


This study was approved by the local ethics committee, and all patients gave written informed consent. 20 people (7 males and 13 females; mean age, 49.1±19.0 years old) were recruited for this study. All MR imaging was performed on a 1.5 T MR scanner (MAGNETOM Aera, Siemens Healthcare, Erlangen, Germany). A prototype sagittal multi-echo Dixon VIBE sequence was used for fat quantification with the following parameters: 6 TEs of 1.12, 2.46, 3.69, 4.92, 6.15, and 7.38 ms, a TR of 9.15 ms, a slice thickness of 4.0 mm, a flip angle of 4°, a matrix size of 160×126, and a field-of-view of 327×261mm2. PDFF was calculated inline using a prototype implementation of a multi-step adaptive fitting approach1. Single-voxel 1H MRS of each L2-4 vertebra was performed using a prototype multi-echo STEAM sequence without water suppression with 5 TEs of 12, 24, 38, 48 and 72 ms; a TR of 3000 ms, and a voxel of 15×15×15 mm3. PDFF was calculated inline using prototype peak fitting of every TE to correct the T2 decay of each component2. Bone mineral density (BMD) of each L3-4 vertebra was measured with single-energy quantitative computed tomography (CT). Linear correlation analysis was performed using the Pearson correlation coefficient to evaluate the association between PDFF measured by Dixon VIBE and MRS with BMD.


L2, L3 and L4 showed no difference in BMD (124.3 mg/cm3±48.3 vs. 131.3 mg/cm3±52.6 vs. 129.9 mg/cm3±54.8; P >0.05), PDFF measured by 1H-MRS (37.5%±8.7 vs. 41.5%±9.4 vs. 42.5%±10.3; P >0.05), and PDFF measured by Dixon VIBE (46.9%±10.7 vs. 50.0%±11.2 vs. 48.6%±11.1; P >0.05). Thus, the values from L2, L3 and L4 were grouped together in the following analysis. PDFF measured by 1H-MRS was lower than that measured by Dixon VIBE (40.5%±9.8 vs. 48.5%±11.2, P <0.001), but had significant correlation (r = 0.747, P <0.001; Figure 1). PDFF measured by 1H-MRS and Dixon VIBE both showed significant correlation with BMD measured by CT (r = -0.717 and -0.712, P <0.001 for both; Figure 2).


This study compared the PDFF measured by 1H-MRS and Dixon VIBE using multiple echoes, and assessed the correlation with BMD. The results showed significant correlation between them. FF measured by single-echo 1H-MRS is TE dependent, and the multi-echo 1H-MRS used in the present study corrected the T2 decay to get a reliable derivation of FF. As 1H-MRS only measured FF in a VOI and needs to be performed in each vertebra, Dixon VIBE has an advantage to cover a larger FOV and get the FF map of all vertebrae in one examination. Using the multi-echo Dixon VIBE, which is insensitive to field inhomogeneity and corrected for T2* decay, the PDFF has significant correlation with PDFF measured by 1H-MRS, and also shows strong correlation with BMD.


Multi-echo MRS and multi-echo Dixon VIBE are useful in assessing spinal fat content, and can be used to assess the BMD.


No acknowledgement found.


1. Zhong X, Nickel MD, Kannengiesser SA, Dale BM, Kiefer B, Bashir MR. Liver fat quantification using a multi-step adaptive fitting approach with multi-echo GRE imaging. Magn Reson Med 2014;72:1353

2. Pineda N, Sharma P, Xu Q, Hu X, Vos M, Martin DR. Measurement of hepatic lipid: high-speed T2-corrected multiecho acquisition at 1H MR spectroscopy--a rapid and accurate technique. Radiology. 2009;252:568


Figure 1. Plot of correlation analysis between PDFF measured by multi-echo 1H-MRS and multi-echo Dixon VIBE in L2, L3 and L4 vertebrae.

Figure 2. Plot of correlation analysis between PDFF measured by 1H-MRS and Dixon VIBE with BMD.

Proc. Intl. Soc. Mag. Reson. Med. 24 (2016)