Validation of a DIXON-based fat quantification technique for the measurement of visceral fat using a CT-based reference standard

Abstract

MR- and CT-based visceral fat quantification are highly correlated and have good cross-modality reliability, indicating that visceral fat quantification by either method can yield a stable and reliable biomarker.

Comparison of CT and Dixon MR Abdominal Adipose Tissue Quantification Using a Unified Computer-Assisted Software Framework

“…Several studies have compared adipose tissue measurements in the abdomen by CT and MR in animals [ 8 ] and humans [ 9 , 10 , 18 , 19 , 20 ]. Their results in general found that both modalities shared a high level of agreement for quantifying abdominal adipose tissue.…”

“…Their results in general found that both modalities shared a high level of agreement for quantifying abdominal adipose tissue. However, one study found a lesser amount of visceral fat in MR compared to CT, possibly attributable to partial volume effects in early MR technology [ 9 ], and another study found a greater amount of visceral fat in MR compared to CT, perhaps attributable to the suboptimal CT thresholds used [ 19 ].…”

“…For example, Kim et al, [ 21 ], Takahashi et al [ 22 ], Parikh et al [ 23 ], and Ozola-Zālīte et al [ 24 ] presented dedicated methods for abdominal fat quantification, but they only worked for CT images. Heckman et al [ 19 ], Positano et al [ 25 ], Liou et al [ 26 ], Zhou et al [ 27 ], Würslin et al [ 28 ], and Kullberg et al [ 29 ] developed different techniques for quantifying abdominal adipose tissues only for the MR images.…”

“…While a HU range from −190 to −30 was more commonly used for identifying adipose tissue in CT [ 20 , 21 , 22 , 23 , 24 ], a HU range from −190 to −50 HU was found to produce excellent reproducibility and repeatability for CT-based abdominal SAT and VAT quantification [ 35 ]. Other HU thresholds have also been used on the CT images—for example, from −150 to −30 and from −205 to −51 to measure abdominal visceral fat [ 19 , 36 ], from −140 to −30 for measuring adipose tissue in a PET/CT system [ 37 ], and from −190 to −40 and from −190 to −45 to measure epicardial adipose tissue [ 38 ]. Therefore, there is no standard CT threshold range for measuring adipose tissue, and these thresholds may need to be adjusted and optimized for different imaging applications.…”

“…For example, Kim et al, [21], Takahashi et al [22], Parikh et al [23], and Ozola-Zālīte et al [24] presented dedicated methods for abdominal fat quantification, but they only worked for CT images. Heckman et al [19], Positano et al [25], Liou et al [26], Zhou et al [27], Würslin et al [28], and Kullberg et al [29] developed different techniques for quantifying abdominal adipose tissues only for the MR images. Several commercial software platforms have been used to quantify abdominal fat in CT and MR scans, but their analysis workflow required manual delineation of the abdominal wall muscle layer.…”