Precision of quantitative ultrasound: comparison of three commercial scanners

Quantitative ultrasound (QUS) measurements of bone have been shown to be independent predictors of osteoporotic fracture risk. Drawbacks of this technique have included the precision of the scanners, which is said to be poorer than in dual-energy X-ray absorptiometry (DXA), in part due to difficulty in repositioning of the foot in an os calcis system and difficulty in comparison across different technologies. A new type of QUS scanner has been introduced that produces an image of the area scanned and is believed to improve precision by aiding repositioning. In this study, we compare three scanners: a dry system (McCue CUBA Clinical); a nonimaging water-bath system (Lunar Achilles+); and an imaging water-bath system (Osteometer DTU-One). Short-term phantom precision was calculated by repeating measurements ten times in succession on the manufacturer-supplied phantom. Long-term phantom precision was calculated by examining the phantom measurements over a 6 month period. In vivo precision was calculated in 26 normal volunteers (19 women, 7 men) and 20 women with osteoporosis. Monitoring time intervals (MTIs) were also calculated using the manufacturer’s normative database. The MTI is the period between scans required to show that a “true” change has occurred, and was between 0.5 year for stiffness (a derived index produced by the Lunar Achilles instrument) and >5 years for all other measurements. The imaging system did not seem to improve precision. Precision for the QUS phantom was similar to that of DXA with a coefficient of variation (CV) of around 1.5% for BUA and <1% for speed of sound (SOS). The precision was such that the technique may be considered for monitoring skeletal changes. However, the change of bone mass at the os calcis in response to treatment was slow, which made the time needed to wait before assessing change, on the whole, longer than that for DXA. An exception may be the Lunar Achilles “stiffness” parameter, but this can only be determined in a longitudinal, comparative treatment study.