Application of ultrasound for quantitative assessment of body fat mass

Obesity is a significant health concern associated with various diseases. Accurate measurement of body fat mass (BFM) and local fat thickness (FT) is crucial for health assessment. Ultrasound offers a non-invasive, portable, and cost-effective alternative for measuring FT, but its application for quantitative BFM estimation has not been fully explored. This study aimed to develop and validate a quantitative estimation algorithm for BFM based on local FT measured by ultrasound.

A total of 179 volunteers were randomly divided into modeling and verification groups. BFM was measured using bioelectrical impedance analysis (BIA), and FT was measured at 10 sites throughout the body using ultrasound. In the modeling group, the correlation between FT and BFM at different sites was analyzed, and a BFM estimation algorithm based on FT was developed using multiple linear regression. The accuracy of the estimation equation was validated in the verification group.

Men had lower BFM than women (P < 0.05). At most sites, the FT of males was less than that of females (P < 0.001). Significant positive correlations were observed between FT at various sites (site 1 to 10) and BFM across all groups (P < 0.01). The estimation algorithm revealed that FT at 4 sites (intra-abdominal, posterior right perinephric, abdominal subcutaneous, and anterior upper arm) contributed to BFM estimation for men, while two additional sites (pre-peritoneal and posterior lower leg) were valuable for women. The R² for the algorithms was 0.882 for men and 0.907 for women, with the standard error of estimate of 2.04 kg for both. The intraclass correlation coefficient between ultrasound-derived estimated BFM and the BFM measured by BIA in the verification group was 0.848 (P < 0.001).

BFM can be quantitatively estimated using a fitting algorithm based on ultrasound-derived local FT.