Improved reproducibility of broadband ultrasound attenuation of the os calcis by using a specific region of interest.

Quantitative ultrasound (QUS) bone measurement is a promising, relatively new technique for the diagnosis of osteoporosis. Contrary to the more established method of bone densitometry (measurement of bone mineral density, BMD, e.g., using dual X-ray absorptiometry, DEXA), QUS does not employ ionizing radiation. It has, however, been a problem to achieve sufficient reproducibility of the QUS measurements. The aim of this study is to evaluate the possible advantages of measuring broadband ultrasound attenuation (BUA) at a region of interest (ROI) instead of at a fixed position, in terms of in vivo precision and correlation to hip bone mineral density. BUA was measured in 27 premenopausal women, 28 postmenopausal women, an 22 men on the DTU-one. Using high resolution images, a ROI is defined in the posterior part of the os calcis as an area with a local minimum of attenuation and a fixed position within the os calcis is defined relative to the water bath. All BUA measurements were performed twice. BMD at the hip was measured on the QDR-2000. The mean BUA values were significantly different between pre- and postmenopausal women, p = 0.0001 for both the ROI (BUAROI) and the fixed position (BUAFIX). The ROI was found in all subjects and was readily reproducible. The precision at the ROI: 1.20 CV% (95% CI: 1.01-1.29 CV%) was significantly better than at the fixed position: 3.87 CV% (95% CI: 3.23-4.48 CV%). BUAROI (r = 0.64) correlated significantly better than BUAFIX (r = 0.35) with HIP BMD. In conclusion, the use of an imaging technique enables BUA measurements to be performed at a ROI. The precision of BUA at the ROI is significantly better than at the fixed position. BUA measured at the ROI correlates better with HIP BMD than BUA measured at the fixed position.