Information Technology Reference
In-Depth Information
Shen et al. [
101
] compared measurements of vertebral bone marrow fat fraction
of L3 and the femoral neck in 27 postmenopausal women as assessed by
T
1
T1-weighted MR imaging, the Dixon method, and
1
H-MRS. Correlations between
the obtained bone marrow fat fractions of the three MR methods ranged from
r = 0.78 to 0.88 in L3. Karampinos et al. [
111
] measured bone marrow fat fraction
at the proximal femur in 7 healthy volunteers by using chemical shift-based water-
fat
1
H-MRS. They reported a signi
cant underestimation of the
fat fraction as determined by a
1
H-MRS model which did not account for short
T
2
* species compared to the chemical shift-based water-fat imaging-based fat
fraction. However, a good equivalency was observed between the fat fraction by
using a
1
H-MRS model accounting for short T
2
* species and the chemical shift-
based water-fat imaging-based fat fraction (R
2
= 0.87). Thus, the comparison of
bone marrow fat fraction derived from different MR methods is limited and
potential confounders have to be taken into account.
Grif
imaging and
L4 and
1
H-MRS of L3
in 82 men with a mean age of 73 years. According to their DXA results, 42 subjects
were classi
th et al. [
102
] performed DXA measurements of L1
-
ed as healthy (mean T-score of 0.8
±
1.1), 23 subjects as osteopenic
(mean T-score of 1.6
±
0.4), and 17 subjects as osteoporotic (mean T-score of
1
H-MRS-based vertebral bone marrow fat content of L3 was signifi-
±
3.2
0.5).
-
±
cantly increased in osteoporotic subjects (58.23
7.8 %) and osteopenic subjects
(55.68
±
10.2 %) compared to subjects with normal BMD (50.45
±
8.7 %). Similar
results were observed in a study population of 103 postmenopausal women older
than 65 years of age [
103
]. Based on DXA measurements of L1
L4, 18 women had
a normal BMD, 30 women were osteopenic, and 55 women had osteoporosis.
Vertebral bone marrow fat content of L3 was signi
-
cantly increased in the osteo-
porotic group (67.8
±
8.5 %) when compared with that of the normal BMD group
(59.2
10.0 %). Similar results were observed for T
1
T1-weighted MRI and chemical
shift-based water-fat imaging. Shen et al. [
105
] obtained T
1
T1-weighted MRI in 210
healthy African-American and Caucasian men and women aged 38
±
52 years. Hip
and lumbar spine BMD were measured by DXA. Pelvic, hip, and lumbar spine
bone marrow adipose tissue as assessed by T
1
T1-weighted MRI showed negative
correlations with hip and lumbar spine BMD (r =
-
0.399 to
0.550). These neg-
−
−
ative relationships remained signi
cant after adjusting for demographics and body
composition. K
ü
hn et al. studied 51 patients (28 female; mean age 69.7
±
9.0 years)
who underwent DXA from L1
L4 to measure BMD and chemical shift-based
water-fat imaging to determine vertebral bone marrow fat fraction [
107
]. The 173
investigated vertebral were divided into three groups (healthy, osteopenic, and
osteoporotic) based on their T-score. The area-under-the-curve to differentiate
between normal and osteoporotic vertebrae was statistically signi
-
cant with
AUC = 0.656. The authors therefore concluded that osteoporosis is associated with
increased vertebral bone marrow fat.
Furthermore,
1
H-MRS allows the calculation of the unsaturation level, previ-
ously de
ned using the formula [
112
,
113
]:
