Influence of body size on the accuracy of lumbar bone mineral density by dual-energy CT: A phantom study

To elucidate the relationship between lumbar bone mineral density (BMD) measurements obtained using dual-energy X-ray absorptiometry (DXA) and dual-energy computed tomography (DECT) and to compare the impact of body size on the precision of BMD measurements obtained using both DXA and DECT.

Abdominal phantoms simulating four body sizes and containing four bone-equivalent phantoms with varying CaCO₃ densities were used. The phantoms were scanned using DXA and DECT at fixed imaging conditions. The relationship between the DXA-measured BMD and the DECT CT value difference (CT value of low tube voltage - CT value of high tube voltage) for the small (S) size phantom was estimated using simple regression analysis. The mean DXA-derived BMD of the S-size phantom was used as the reference value. The effect of four body sizes on the precision of DECT and DXA BMD was measured using the regression equation and compared with the acceptable range of 5.3% least significant change (LSC).

The coefficient of determination of simple regression analysis between DXA-measured BMD and DECT CT value difference demonstrated a remarkably strong linear relationship (R² = 0.993). DXA accurately measured BMD up to body size medium within 5.3% LSC but tended to underestimate it for body sizes large (L) and double large (LL). On the other hand, DECT accurately measured BMD up to body size L within 5.3% LSC but tended to overestimate BMD at body size LL.

DXA and DECT BMD measurements exhibited a strong linear correlation. Body size had a lower influence on the accuracy of BMD measurement obtained using DECT than using DXA, demonstrating higher precision even for larger body sizes.