Bolotin H H
School of Physics, University of Melbourne, Parkville, Victoria, Australia.
Med Phys. 1998 Feb;25(2):139-51. doi: 10.1118/1.598175.
An extensive analytic exposition is developed of systematic inaccuracies inherent in planar dual-energy x-ray absorptiometry (DXA) in vivo bone mineral density (BMD) measurements arising directly from the dual-energy facet of this methodology. Generalized and specific criteria governing these BMD inaccuracies are derived for the presence of relevant absorptiometrically distinguishable extra- and intra-osseous soft tissues (lean muscle tissue, interposed adipose, marrow, and cerebrospinal fluid). These analytic findings are utilized in comprehensive quantitative simulation studies which evaluate the magnitude and stipulate the direction (under- or over-estimate) of such systematic errors for typical, realistic DXA in vivo lumbar vertebral BMD scans over the ranges of soft tissue parameters and trabecular bone volumes (TBV) encountered clinically. It is shown that inherent systematic inaccuracies as high as +/- approximately 20% may be anticipated in typical patient-specific planar DXA vertebral BMD scans, particularly so for older, osteoporotic-prone, and osteoporotic individuals. These inaccuracies exceed considerably the DXA precision error and may mask or exaggerate clinically significant true changes in BMD. Both the scale and trends of patient-specific inaccuracies found to pertain in the present work are in keeping with those observed by recent investigators in actual specimen-specific in situ and in vitro DXA scans of given cadaveric bones. Insofar as the range of BMD inaccuracies found here is comparable to that of variations in the DXA-measured interpatient, age-moderated BMD of normal individuals, the intrinsic specificity of planar DXA for screening/diagnostic purposes is examined critically. It is shown that these inherent inaccuracies may prompt erroneous diagnoses, assessments, and interpretations of DXA-derived BMD measurements undertaken to screen, monitor, and help evaluate patient-specific predictive bone fragility, and to assess the efficacy of drug and other therapeutic regimens.
本文对平面双能X线吸收法（DXA）体内骨密度（BMD）测量中固有的系统误差进行了广泛的分析阐述，这些误差直接源于该方法的双能特性。针对存在相关吸收测量可区分的骨外和骨内软组织（瘦肌肉组织、中间脂肪、骨髓和脑脊液）的情况，推导了控制这些BMD误差的通用和特定标准。这些分析结果用于全面的定量模拟研究，评估临床遇到的软组织参数和小梁骨体积（TBV）范围内典型、实际的DXA体内腰椎BMD扫描中此类系统误差的大小，并规定其方向（低估或高估）。结果表明，在典型的患者特异性平面DXA椎体BMD扫描中，可能预期存在高达±约20%的固有系统误差，对于老年、易患骨质疏松症和骨质疏松症患者尤其如此。这些误差大大超过了DXA的精确误差，可能掩盖或夸大BMD临床上的显著真实变化。在本研究中发现的患者特异性误差的规模和趋势与最近研究人员在给定尸体骨骼的实际标本特异性原位和体外DXA扫描中观察到的一致。鉴于此处发现的BMD误差范围与DXA测量的正常个体患者间、年龄调节的BMD变化范围相当，对平面DXA用于筛查/诊断目的的内在特异性进行了严格审查。结果表明，这些固有误差可能导致对为筛查、监测和帮助评估患者特异性预测性骨脆性以及评估药物和其他治疗方案疗效而进行的DXA衍生BMD测量的错误诊断、评估和解释。
