Estimating proton density fat fraction using computed tomography attenuation in skeletal muscle and adipose tissue in a prospective clinical cohort.
Authors
Affiliations (6)
Affiliations (6)
- Department of Nutritional Sciences, University of Wisconsin, Madison, WI, USA.
- Department of Radiology, University of Wisconsin, Madison, WI, USA; Department of Medical Physics, University of Wisconsin-Madison, Madison, WI, USA; Department of Biomedical Engineering, University of Wisconsin-Madison, Madison, WI, USA.
- Department of Radiology, University of Wisconsin, Madison, WI, USA; Department of Medical Physics, University of Wisconsin-Madison, Madison, WI, USA.
- Department of Radiology, University of Wisconsin, Madison, WI, USA; Department of Medical Physics, University of Wisconsin-Madison, Madison, WI, USA; Department of Biomedical Engineering, University of Wisconsin-Madison, Madison, WI, USA; Department of Medicine, Emergency Medicine, University of Wisconsin, Madison, WI, USA.
- Department of Radiology, University of Wisconsin, Madison, WI, USA.
- Department of Nutritional Sciences, University of Wisconsin, Madison, WI, USA. Electronic address: [email protected].
Abstract
Proton density fat fraction (PDFF) measured using magnetic resonance imaging (MRI) is considered a non-invasive reference measure of fat deposition in various tissues and has been proposed as a quantitative indicator of tissue quality. While the contraindications of MRI limit PDFF as a routine clinical tool, computed tomography (CT) scans are often available clinically. Currently, there are no standard methods to compare CT and MRI tissue density measures. The objective of this study was to determine the relationship between CT and MRI tissue density values in skeletal muscle and subcutaneous adipose tissue (SAT). Tissue density (using same-day CT and MRI) was assessed for both muscle and SAT in a healthy prospective clinical cohort (n = 50, 23 males, 27 females; mean age 57 ± 5 years; body mass index 27 ± 5). Comparisons using small regions of interest and entire cross-sectional areas were made at both L1 and L3 vertebral levels. CT cross-sectional area was measured using automated AI-based segmentation software. All other measurements were demarcated manually. Regressions with 95% confidence intervals were fitted separately for each comparison. A strong negative linear correlation was found for all CT Hounsfield Unit (HU)-PDFF tissue density pairings. The strongest association for skeletal muscle tissue density was measured at L3 using cross-sectional area (PDFF (%) = -0.352 × mean CT attenuation (HU) + 26.4; r<sup>2</sup> = 0.791). The strongest linear correlation for SAT density was identified at L1 using cross-sectional area (PDFF (%) = -0.604 × mean CT attenuation (HU) + 31.2; r<sup>2</sup> = 0.885). It is feasible to convert density values from CT scans to accurate estimates of PDFF in skeletal muscle and SAT. Conversion of tissue density values from CT HU to estimated PDFF expand clinical measures of muscle quality and improve body composition assessment.