Sound-tissue interaction: the physical basis of bone ultrasonometry and limitations of existing methods.

Ultrasound measurements of bone are generally obtained using transmission rather than pulse-echo techniques because of its highly attenuating nature. Ultrasound velocity and attenuation measurements are utilized. For velocity, there are well-defined fundamental relationships describing the dependence on the elasticity and density of bone. However, the practical implementation and signal processing of velocity measurements has led to a significant variability in results from different commercial systems. We may measure either phase of group velocity, for the latter, adopting a range of pulse arrival definitions. We are offered bone velocity, heel velocity, time of flight, and amplitude-dependent velocity. For attenuation measurements, however, the reverse is true. We generally record the increase in attenuation with frequency (0.2-0.6 MHz), termed broadband ultrasound attenuation (BUA). Although first described in 1984, because of the complex interplay of attenuation mechanisms, there still lacks a fundamental understanding of the dependence of BUA on the material and structural properties of cancellous bone. With the increasing number of commercial systems available, there is an urgent need to understand the intrinsic (artefact free) and system estimation of ultrasound velocity and attenuation parameters that may be implemented to characterise bone and provide clinical information.