Standardization of CT densitometry.
Standardization of CT densitometry.
复制标题
CT 密度测定的标准化。
作者：
D. Parr;R. Stockley
编辑：定量计算机断层扫描（CT）技术的发展，用于测量慢性肺部疾病，如肺气肿和特发性肺纤维化是一个重大的进步，在追求新的疾病修饰疗法，但中心之间的标准化仍然是一个重大障碍。我们饶有兴趣地阅读了Best博士及其同事在2003年8月的《放射学》杂志上发表的文章（1），其中将校正因子应用于肺密度测量，以实现中心之间的CT扫描仪可比性。以前有人建议，校正CT扫描仪之间的空气校准差异可能会实现标准化（2），但我们认为这项最新研究中采用的技术可能不是最佳方法。感兴趣区域的肺内空气测量具有高度的空间依赖性（3），需要采用严格的采样技术以确保重现性。此外，这些测量受患者体质的影响，我们的研究小组发现体重指数和气管空气密度测定之间具有良好的相关性（r 0.55，P 0.001）（4）。这两个因素可能会引入变异性，并可能部分解释为什么在没有“校正”的情况下，平均肺衰减和生理变量之间的相关性略高于上级。作者尚未确认校正技术，也未评论其是否成功地验证了使用不同CT扫描仪获得的测量值。在一项对127名患者进行的研究中，通过使用校准良好的CT扫描仪，我们发现气管空气密度的测量值与患者体外获得的空气测量值（平均差异为22.5 HU 10.89 [SD]）和预期值1，000 HU存在很大差异。在这种技术被广泛接受之前，应更充分地探讨其原因以及将这种测量结果纳入校正因子对密度测定的影响。
Editor: The development of quantitative computed tomographic (CT) techniques for the measurement of chronic lung diseases such as emphysema and idiopathic pulmonary fibrosis is a major advance in the pursuit of novel disease-modifying therapies, but standardization between centers remains a substantial obstacle. We read with interest the article by Dr Best and colleagues in the August 2003 issue of Radiology (1) in which a correction factor was applied to lung density measurements to achieve CT scanner comparability between centers. It has previously been suggested that correcting for differences in air calibration between CT scanners may allow standardization (2), but we believe the technique employed in this latest study may not be the optimum method. Region of interest measurements of intratracheal air are highly spatially dependent (3) and necessitate a rigorous sampling technique to ensure reproducibility. In addition, such measurements are influenced by patient habitus, and our group has found good correlation between body mass index and tracheal air densitometry (r 0.55, P .001) (4). These two factors are likely to introduce variability and may partly explain why correlation between mean lung attenuation and physiologic variables is marginally superior in the absence of “correction.” The authors have not validated the correction technique and do not comment on whether it successfully standardizes measurements obtained with different CT scanners. By using a well-calibrated CT scanner in a study of 127 patients, we found that measurements of tracheal air density gave widely different values from air measurements obtained external to the patient (mean difference, 22.5 HU 10.89 [SD]) and from the intended value of 1,000 HU. The reasons for this and the influence that correction factors incorporating such measurements impose on densitometry should be explored more fully before such techniques are widely accepted.