Soft-tissue thickness compensation for ultrasound transit time spectroscopy estimated bone volume fraction—an experimental replication study

Abstract

Quantitative ultrasound offers a reliable means to predict osteoporotic fracture risk, although to date it has not been generally used within routine clinical management since it does not provide a direct estimate of bone mineral density (BMD), and hence, the associated WHO criteria for osteopenia and osteoporosis. Langton has proposed that ultrasound propagation through cancellous bone may be considered as an array of parallel sonic-rays, the transit time of each determined by the corresponding proportion of bone and marrow propagated. This concept has led to the development of ultrasound transit time spectroscopy (UTTS) to estimate solid (bone) volume fraction (SVF). However, within the real-world clinical environment, a bone, such as the calcaneus, has overlying soft-tissues that would result in a significantly time-extended transit time spectrum (TTS), and hence, an underestimated SVF. The aims of this experimental replication study were firstly, to investigate the effect of overlying soft-tissues upon UTTS derived SVF (UTTS-SVF) estimation, and secondly, to develop and evaluate a method to compensate for this, thereby providing a more accurate estimation of SVF. Four 3D-cylindrical replica cancellous bone samples, with flat-parallel cortex discs on opposite faces, were studied; with varying thicknesses of water-replicating overlying soft-tissues. Through-transmission ultrasound signals were recorded, from which the apparent TTS was derived via deconvolution. Pulse-echo signals were utilised to measure the thickness of water overlying the replica cortices. The TTS was then corrected for the ultrasound transit time associated with the overlying water. UTTS-SVF was then calculated, and compared with the SVF value measured with microcomputed tomography (μCT-SVF). The results demonstrated that varying water-thicknesses for each sample provided very similar formats of ultrasound transit time spectra, but with significant extended time shifts. Compensation for overlying water thickness provided an accurate estimate of SVF for all samples; the overall of agreement between UTTS-SVF with μCT-SVF being 92.68%. It is therefore suggested that UTTS has the potential to provide a reliable in vivo estimate of BMD and hence application of the established WHO T-score for routine clinical assessment of osteoporosis.