Abstract
Purpose: To evaluate the diagnostic performance of magnetic resonance imaging-estimated proton density fat fraction (MRI-PDFF) using a free-drawn region-of-interest (ROI) measurement of hepatic fat deposition compared to MR spectroscopy (MRS) as the reference standard. Materials and methods: A total of 156 patients underwent 3T MR imaging with a multi-step adaptive fitting approach, multi-echo volume interpolated breath-hold examination (VIBE) acquisition and single-voxel high-speed T2-corrected multiple-echo 1H-MR spectroscopy (SVS). Seven ROI measurements were performed in each segment of the fat percentage maps (“segmental-PDFF”). Three ROIs were placed at the same level as the SVS (“VOI-PDFF”). Free-hand-drawn ROIs were placed at three different levels along the entire liver (“free-drawn-PDFF”) and separately along the right and left lobes (“free-drawn-PDFF-2”). Results: A strong correlation was found between VOI-PDFF and SVS (r = 0.977). The right lobe had greater fat content than the left lobe (p < 0.001). After image analysis, 54 and 46 patients were classified as having steatosis using SVS and free-drawn-PDFF as a reference standard, respectively. The diagnostic performance of free-drawn-PDFF was significantly different from SVS (p < 0.05). Conclusion: Free-drawn-PDFF provides accurate and generalized information regarding hepatic fat deposition. It is a useful method, particularly if fat deposition is heterogeneous, and should be considered as a new reference standard.
| Original language | English |
|---|---|
| Pages (from-to) | 1083-1093 |
| Number of pages | 11 |
| Journal | Radiologia Medica |
| Volume | 120 |
| Issue number | 12 |
| DOIs | |
| State | Published - 2015.12.1 |
Keywords
- Fat quantification
- Hepatic steatosis
- MR spectroscopy
- Proton density fat fraction
Quacquarelli Symonds(QS) Subject Topics
- Medicine
Fingerprint
Dive into the research topics of 'Hepatic fat quantification using the proton density fat fraction (PDFF): utility of free-drawn-PDFF with a large coverage area'. Together they form a unique fingerprint.Cite this
- APA
- Author
- BIBTEX
- Harvard
- Standard
- RIS
- Vancouver