Kobayashi T
Department of Radiology, Kanazawa University, School of Medicine.
Nihon Igaku Hoshasen Gakkai Zasshi. 1995 Aug;55(9):638-45.
Bone images obtained by single exposure dual energy subtraction with computed radiography, a new image processing method, were evaluated for the detectability of calcification in lung nodules in both phantom and clinical studies. In the phantom study, five simulated nodules containing known and uniformly distributed amounts of calcium hydroxide were placed in a human chest phantom and images were obtained the following five techniques: 1) conventional radiography (CXR), 2) low kV conventional radiography (Low), 3) Fuji computed radiography (FCR), 4) bone images with conventional singles exposure energy subtraction (ES) and 5) bone image with fine single exposure energy subtraction (FES). The detectability of calcification in the simulated nodules was evaluated by six radiologists using ROC analysis and two-way analysis of variance. In the clinical study, images of 42 lung nodules (21 calcified lung nodules and 21 non-calcified nodules), in which calcification was diagnosed by computed tomography, were obtained with the following four techniques: 1) CXR, 2) FCR, 3) ES and 4) FES and the detectability of calcification in the lung nodules was evaluated by nine radiologists using ROC analysis. In the phantom study, FES was significantly superior to CXR, Low and ES, and equal to FCR in the detectability of calcification in the simulated nodules. Compared with ES, FES was especially superior in the underexposed portion of the chest radiograph. In terms of calcium content in the simulated nodules, FES and FCR were superior to CXR in detecting calcification when a lower calcium hydroxide content was present. In the clinical study, FES was significantly superior to CXR, FCR and ES in the diagnosis of calcification in lung nodules. These findings indicate the usefulness of FES in the detection of calcification in lung nodules, and suggests that it is promising for the diagnostic screening of lung nodules.
采用一种新的图像处理方法——计算机X线摄影单次曝光双能减影获得的骨图像，在体模和临床研究中对肺结节钙化的可检测性进行了评估。在体模研究中，将五个含有已知且均匀分布氢氧化钙量的模拟结节置于人体胸部体模中，并采用以下五种技术获取图像：1）传统X线摄影（CXR），2）低千伏传统X线摄影（Low），3）富士计算机X线摄影（FCR），4）传统单次曝光能量减影骨图像（ES）和5）精细单次曝光能量减影骨图像（FES）。六名放射科医生使用ROC分析和双向方差分析对模拟结节中钙化的可检测性进行了评估。在临床研究中，对42个肺结节（21个钙化肺结节和21个非钙化结节）的图像进行了获取，这些结节的钙化通过计算机断层扫描诊断，采用以下四种技术：1）CXR，2）FCR，3）ES和4）FES，九名放射科医生使用ROC分析对肺结节中钙化的可检测性进行了评估。在体模研究中，FES在模拟结节钙化的可检测性方面显著优于CXR、Low和ES，与FCR相当。与ES相比，FES在胸部X光片曝光不足的部分尤其优越。就模拟结节中的钙含量而言，当氢氧化钙含量较低时，FES和FCR在检测钙化方面优于CXR。在临床研究中，FES在肺结节钙化的诊断方面显著优于CXR、FCR和ES。这些发现表明FES在检测肺结节钙化方面的有用性，并表明其在肺结节诊断筛查方面具有前景。
