DOI:10.2214/AJR.07.2992
AJR 2007; 189:W385
© American Roentgen Ray Society
Reply
Darren D. Brennan,
Paul F. Whelan,
Kevin Robinson,
Ovidiu Ghita,
Julie M. O'Brien,
Robert Sadleir and
Stephen J. Eustace
Beth Israel Deaconess Medical Center, Boston, MA
WEB—This is a Web exclusive article.
We appreciate the comments made by Drs. Mantatzis and Prassopoulos
[1] and the opportunity to
reply. In our article [2], we
used a histogram-matching technique to correct for mismatches in signal
intensity between different stacks of images before applying a four-point
segmentation algorithm to segment out fat from whole-body MR images. Mantatzis
and Prassopoulos' are correct in pointing out that this led to the inclusion
of bone marrow adipose tissue. Although they are correct in pointing out that
bone marrow adipose tissue has not been associated with the metabolic syndrome
as has for instance visceral fat
[3], we do not believe that
inclusion of bone marrow adipose tissue invalidates our system and indeed, the
quantitative assessment of marrow fat as a predictor of osteoporosis holds
potential [4].
We do agree, however, that the addition of segmentation could be extremely
useful—particularly to quantify visceral body fat distribution as has
been proposed by recent authors
[5]. Semiautomated manual
segmentation can be easily achieved after simple thresholding by identifying a
seed point in the area of interest and then using a region-growing algorithm
so that all adjacent fat is included. At the risk of introducing a slight
underestimate in the degree of fat measured, the technique can be made easier
by acquiring out-of-phase images. In these, phase-cancellation india-ink
artifact at the boundaries of tissues with bulk fat will occur, thus allowing
accurate delineation of the boundaries of bulk fat tissue. The results of a
recent study show that single-level slice results and multilevel slice results
for measurement of visceral fat yield nearly identical results
[5], which means that such
proposed segmentation techniques could be rapidly performed.
In relationship to bone marrow, segmentation could be used to quantify the
total distribution of bone marrow adipose tissue that could be subtracted from
total body fat, but perhaps more important, segmentation might be used as a
noninvasive tool for the determination of osteoporosis. The ability to
accurately determine the distribution of bone-marrow adipose tissue might even
hold potential for the determination of future fracture risk.
References
- Mantatzis M, Prassopoulos P. Total body fat, visceral fat, bone
marrow fat? What is important to measure? (letter).
AJR 2007;189
:Wxxx
- Brennan DD, Whelan PF, Robinson K, Ghita O, O'Brien JM, Sadleir R,
Eustace SJ. Rapid automated measurement of body fat distribution from
whole-body MRI. AJR 2005;185
: 418-423[Abstract/Free Full Text]
- Snijder MB, van Dam RM, Visser M, Seidell JC. What aspects of body
fat are particularly hazardous and how do we measure them? Int J
Epidemiol 2006; 35:83
-92[Free Full Text]
- Griffith JF, Yeung DK, Antonio GE, et al. Vertebral marrow fat
content and diffusion and perfusion indexes in women with varying bone
density: MR evaluation. Radiology 2006;241
: 831-838[Abstract/Free Full Text]
- Siegel MJ, Hildebolt CF, Bae KT, Hong C, White NH. Total and
intraabdominal fat distribution in preadolescents and adolescents: measurement
with MR imaging. Radiology 2007;242
: 846-856[Abstract/Free Full Text]
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
Top
References
