双能量计算机断层扫描用于评估直径≤3 cm的肺结节强化情况

Dual-energy Computed Tomography for the Evaluation of Enhancement of Pulmonary Nodules≤3 cm in Size.

作者信息

Reiter Michael J, Winkler Weston T, Kagy Kenton E, Schwope Ryan B, Lisanti Christopher J

机构信息

*Department of Radiology, Stony Brook University Medical Center, Stony Brook, NY †Department of Radiology, Brooke Army Medical Center, San Antonio, TX ‡Uniformed Services University of the Health Sciences, Bethesda, MD.

出版信息

J Thorac Imaging. 2017 May;32(3):189-197. doi: 10.1097/RTI.0000000000000263.

DOI:10.1097/RTI.0000000000000263

PMID:28338536

Abstract

PURPOSE

The aim of the study was to compare the accuracies of 4 different methods of assessing pulmonary nodule enhancement to distinguish benign from malignant solid pulmonary nodules using nondynamic contrast-enhanced dual-energy computed tomography.

MATERIALS AND METHODS

Seventy-two patients (mean age, 62 y) underwent dual-energy chest computed tomography 3 minutes after intravenous contrast administration. Each of 118 pulmonary nodules (9±5.9 mm) were evaluated for enhancement by 4 methods: visual assessment, 3-dimensional automated postprocessing measurement tool, manually drawn region of interest with calculated iodine-related attenuation, and measurement of iodine concentration. The optimal cutoff for enhancement was defined as having the largest specificity among all cutoffs while maintaining 100% sensitivity. Accuracy of the methods was assessed with receiver operating characteristic curves.

RESULTS

Ninety-three of 118 pulmonary nodules were benign (79%). Visual assessment of enhancement had sensitivity and specificity of 100% and 44%, respectively. For the automated 3-dimensional measurement tool, 20 HU was found to be the optimal threshold for defining enhancement, resulting in a specificity of 71% and a sensitivity of 100%, as well as an area under the curve (AUC) of 0.87 (95% confidence interval [CI], 0.82-0.92). The AUC was 0.79 (95% CI, 0.73-0.85) for the measured enhancement using a manually drawn region of interest. When a threshold of 21 HU was used for defining enhancement, maximum specificity was obtained (56%) while maintaining 100% sensitivity. The AUC for measured iodine concentration was 0.79 (95% CI, 0.77-0.85). At a cutoff iodine concentration of 0.6 mg/mL, the sensitivity was 100% with a specificity of 57%.

CONCLUSIONS

Although use of automated postprocessing had the highest specificity while maintaining 100% sensitivity, there were only minor clinically relevant differences between measurement techniques given that no single technique misclassified a malignant nodule as nonenhancing.

摘要

目的

本研究旨在比较4种不同的评估肺结节强化的方法，使用非动态对比增强双能计算机断层扫描来区分良性与恶性实性肺结节的准确性。

材料与方法

72例患者（平均年龄62岁）在静脉注射对比剂3分钟后接受双能胸部计算机断层扫描。118个肺结节（9±5.9mm）中的每一个都通过4种方法评估强化情况：视觉评估、三维自动后处理测量工具、手动绘制感兴趣区域并计算碘相关衰减以及测量碘浓度。强化的最佳截断值被定义为在所有截断值中具有最大特异性同时保持100%敏感性。使用受试者工作特征曲线评估这些方法的准确性。

结果

118个肺结节中有93个是良性的（79%）。强化的视觉评估敏感性和特异性分别为100%和44%。对于自动三维测量工具，发现20HU是定义强化的最佳阈值，特异性为71%，敏感性为100%，曲线下面积（AUC）为0.87（95%置信区间[CI]，0.82 - 0.92）。使用手动绘制感兴趣区域测量强化的AUC为0.79（95%CI，0.73 - 0.85）。当使用21HU的阈值定义强化时，获得了最大特异性（56%），同时保持100%敏感性。测量碘浓度的AUC为0.79（95%CI，0.77 - 0.85）。在碘浓度截断值为0.6mg/mL时，敏感性为ioo%，特异性为57%。