INTRODUCTION

SONIALVISION G4 (Shimadzu, Kyoto, Japan) is a fluoroscopic imaging system capable of dual-energy X-ray absorptiometry (DXA) using the additional application. However, bone mineral density (BMD) with this system is calibrated only by a phantom, and the necessity to confirm the reliability of BMD measurement in human subjects has been pointed out. This study aimed to evaluate the clinical reliability of BMD measurement using SONIALVISION G4 by comparing it with PRODIGY (GE Healthcare, Madison, WI, USA), a reference system in developing the DXA application.

MATERIALS AND METHODS

BMD measurements using these two DXA systems were performed on the same day at the lumbar spine and bilateral proximal femur of 130 subjects aged 22-86 years. The agreement and correlation between the measurements were determined. Thirty out of the 130 subjects were scanned twice to determine the precision of the SONIALVISION G4 system.

RESULTS

BMD was highly correlated between the two DXA systems both at the lumbar spine and proximal femur, with correlation coefficients ranging from 0.963 (left total hip) to 0.989 (left femoral neck). Differences in BMD measured by the two systems were less than 3.0%, but a small proportional bias was observed in the Bland-Altman analysis. Coefficients of variation in BMD measurement were 1.77% in the lumbar spine and 1.41-2.76% in the proximal femur.

CONCLUSIONS

Although BMD values by SONIALVISION G4 were close to those by PRODIGY, follow-up studies should be performed using the same device because of the small difference between devices.