Effect of Region of Interest Location on Ultrasound Measurements of the Calcaneus

Abstract

Ultrasound (US) measurements of the calcaneus are usually carried out in a region of interest (ROI) at a fixed site relative to a footplate. Recently, US transmission systems have been developed with imaging capability that enable selection of the position of ROI; the region of measurement is always the area of minimum attenuation in the posterior part of the calcaneus. This study compares measurements of broadband ultrasound attenuation (BUA) and speed of sound (SOS) at the variable ROI of minimum attenuation (ROIv) and at fixed coordinates (ROIf). Ultrasound variables were estimated at ROIv and ROIf in 212 female subjects, including 26 patients with osteoporotic fractures. Among the 186 women without fractures, 63 were classified as having osteoporosis on the basis of their vertebral bone density. Precision of BUA and SOS were better at ROIv than at ROIf. BUA was more highly correlated with bone mineral density (BMD) at the lumbar spine and femoral neck at ROIv than ROIf (r = 0.64 for lumbar spine and 0.68 for femoral neck at ROIv versus 0.50 for lumbar spine and 0.54 for femoral neck at ROIf, P < 0.05 for both comparisons). There were no significant differences between the correlations of SOS with axial BMD at ROIv compared with ROIf. Significant difference was found between the areas under the ROC curve for each ultrasound variable at ROIv and ROIf for both groups of patients, subjects with osteoporosis (area under curve = 0.87 for BUA at ROIv versus 0.82 at ROIf, P < 0.05; area under curve = 0.85 for SOS at ROIv versus 0.81 at ROIf, P < 0. 05), and women with fractures (area under curve = 0.93 for BUA at ROIv versus 0.86 at ROIf, P < 0.05; area under curve = 0.91 for SOS at ROIv versus 0.82 at ROIf, P < 0.05). Ultrasound variables measured at ROIv enable improved reproducibility and significantly better differentiation of diseased subjects from healthy individuals as compared with measurements at ROIf.

Does Quantitative Ultrasound Imaging Enhance Precision and Discrimination?

“…These correlations were comparable to those reported in the literature for both non-imaging devices [28][29][30] and imaging devices [19,20]. The correlation with BMD was not significantly higher when the DTU was used despite a significant improvement in BUA short-term reproducibility.…”

“…The short-term precision errors observed in the present study for the DTU are comparable to those reported in the literature for imaging QUS [12,16,19,20]. It seems that, at least in the short term, QUS imaging overcomes some of the limitations of the precision of a measurement taken at a fixed location.…”

“…Large prospective studies utilizing non-imaging devices found that there was a doubling of risk of fracture associated with a 1 SD decline in QUS measurement parameters [2][3][4], and this was comparable to BMD measurements [2,3]. Crosssectional studies on imaging devices have shown that they can discriminate between women with fracture and those without, and in one case, a variable ROI displayed significantly better discrimination compared with a fixed ROI [16,19].…”

“…Therefore, it is important that the spatial inhomogeneity of bone density and structure in the calcaneus can be visualized and an appropriate measurement ROI selected accordingly. Several studies have reported an improvement in precision when an anatomically chosen ROI was used instead of a fixed ROI relative to the footplate when measurements were performed on the same device [19,20]. An automatic ROI can be selected by using anatomic landmarks utilizing image segmentation or using acoustic features such as a region of minimum attenuation.…”

Evaluation of a gel‐coupled quantitative ultrasound device for bone status assessment.

“…It has been suggested that having a variable transducer position might reduce this error. 12 Medilink has adopted this approach in the Osteospace.…”