Adipose tissue MRI for quantitative measurement of central obesity

Abstract

Purpose:
To validate adipose tissue magnetic resonance imaging (atMRI) for rapid, quantitative volumetry of visceral adipose tissue (VAT) and total adipose tissue (TAT).

Materials and Methods:
Data were acquired on normal adults and clinically overweight girls with Institutional Review Board (IRB) approval/parental consent using sagittal 6-echo 3D-spoiled gradient-echo (SPGR) (26-sec single-breath-hold) at 3T. Fat-fraction images were reconstructed with quantitative corrections, permitting measurement of a physiologically based fat-fraction threshold in normals to identify adipose tissue, for automated measurement of TAT, and semiautomated measurement of VAT. TAT accuracy was validated using oil phantoms and in vivo TAT/VAT measurements validated with manual segmentation. Group comparisons were performed between normals and overweight girls using TAT, VAT, VAT-TAT-ratio (VTR), body-mass-index (BMI), waist circumference, and waist-hip-ratio (WHR).

Results:
Oil phantom measurements were highly accurate (<3% error). The measured adipose fat-fraction threshold was 96% ± 2%. VAT and TAT correlated strongly with manual segmentation (normals r2 ≥ 0.96, overweight girls r2 ≥ 0.99). VAT segmentation required 30 ± 11 minutes/subject (14 ± 5 sec/slice) using atMRI, versus 216 ± 73 minutes/subject (99 ± 31 sec/slice) manually. Group discrimination was significant using WHR (P < 0.001) and VTR (P = 0.004).

Conclusion:
The atMRI technique permits rapid, accurate measurements of TAT, VAT, and VTR