Phantoms for evaluating the performance characteristics of bone densitometers

Osteoporosis is characterized by a gradual loss of osseous tissue within the framework of the bone, which lowers bone strength and so increases the risk of fracture. A variety of techniques exist for identifying early bone loss, such as single- or dual-photon absorptiometry (SPA, DP A), quantitative computed tomography (QCT), etc. (Speller et al, 1989). Those commonly employed concentrate on the rate of decrease in bone density or bone mineral content within a particular site. The skeletal sites that become most susceptible to fracture are those with high trabecular content, namely the distal radius, the femoral neck and the lumbar vertebrae. For diagnosis of osteoporosis serial density measurements need to be performed at regular intervals over a period of many months. The frequency at which measurements are performed must be determined by the precision of the machine as compared with the decrease in bone density over that period of time. The reported rates of bone loss for normal post-menopausal women are 1.2% per year for trabecular bone (vertebral QCT measurements; Cann et al, 1985) and 0.9% per year for cortical bone (SPA measurements of mid-radius; Mazess, 1982). Therefore, the precision of the method must be sufficient to allow the machine to detect the small changes in bone density that occur in the space of a few months. Thus, knowledge of the precision and accuracy of the machine is essential. All systems when installed are supplied with a calibration phantom.