Background

Existing CT emphysema measurements quantify the extent or clustering of emphysema voxels in chronic obstructive pulmonary disease (COPD); however, these measurements do not quantify how those voxels are clustered.

Purpose

To develop a CT measurement to quantify the “compactness” of emphysema voxels, called the normalized join count (NJC), and to determine whether the NJC measurement differentiates COPD disease severity and correlates with lung function and visual emphysema scores.

Materials and Methods

In this secondary analysis of a prospective study, lung function and CT images were obtained from the Canadian Cohort Obstructive Lung Disease study visit 1 from 2009 to 2013. Participants were categorized as never-smokers, at risk, mild COPD, or moderate-severe COPD. Diffusion capacity for carbon monoxide/alveolar volume was measured. CT emphysema was scored visually by radiologists. CT measurements included the percentage low-attenuation area with attenuation less than –950 HU (%LAA_–950insp), low-attenuation cluster (LAC), and lowest 15th percentile point of the CT lung density histogram. NJC was developed to measure compactness of CT emphysema voxels. An analysis of variance determined differences between groups. Multivariable ridge regression determined association between CT measurements with lung function and radiologist scores.

Results

A total of 1294 participants (750 men; mean age, 67 years ± 10) were analyzed (277 never-smokers, 306 at risk, 427 mild COPD, and 284 moderate-severe COPD). NJC, %LAA_–950insp, and LAC measurements were higher in moderate-severe COPD than in never-smokers and at-risk participants (P < .05 for all comparisons), but only NJC was different between mild and ;moderate-severe COPD (mean, 1.98% ± 3.61 vs 1.44% ± 2.14; P < .05). In multivariable regression analysis, among all CT measurements NJC had the greatest relative contribution to diffusion capacity for carbon monoxide/alveolar volume (P = .002) and visual emphysema score (P < .001).

Conclusion

The relationship of normalized join count with severity of chronic obstructive pulmonary disease may indicate that the assessment of this disease is dependent on the number of low attenuating voxels or the size of clusters and the spatial arrangement of such voxels.

© RSNA, 2021

Online supplemental material is available for this article.

See also the editorial by Grenier in this issue.

中文翻译：

---

慢性阻塞性肺疾病 CT 肺气肿的空间依赖性通过连接计数统计量化

背景

现有的 CT 肺气肿测量可量化慢性阻塞性肺疾病 (COPD) 中肺气肿体素的范围或聚集；然而，这些测量并没有量化这些体素是如何聚类的。

目的

开发 CT 测量以量化肺气肿体素的“紧密度”，称为标准化连接计数 (NJC)，并确定 NJC 测量是否区分 COPD 疾病严重程度并与肺功能和视觉肺气肿评分相关。

材料和方法

在这项前瞻性研究的二次分析中，从 2009 年至 2013 年加拿大队列阻塞性肺病研究访问 1 中获得了肺功能和 CT 图像。参与者被归类为从不吸烟、有风险、轻度 COPD 或中重度 COPD . 测量一氧化碳/肺泡体积的扩散能力。CT肺气肿由放射科医生视觉评分。CT 测量包括衰减小于 –950 HU (%LAA _–950insp)、低衰减簇 (LAC) 和 CT 肺密度直方图的最低 15% 点。NJC 被开发用于测量 CT 肺气肿体素的紧密度。方差分析确定了组之间的差异。多变量岭回归确定 CT 测量与肺功能和放射科医师评分之间的关​​联。

结果

总共分析了 1294 名参与者（750 名男性；平均年龄，67 岁 ± 10 岁）（277 名从不吸烟，306 名有风险，427 名轻度 COPD 和 284 名中重度 COPD）。NJC、%LAA _–950insp和 LAC 测量值在中重度 COPD 中高于从不吸烟者和有风险的参与者（所有比较P < .05），但只有 NJC 在轻度和中重度 COPD 之间存在差异（平均值，1.98% ± 3.61 对 1.44% ± 2.14；P < .05）。在多变量回归分析中，在所有 CT 测量值中，NJC 对一氧化碳/肺泡体积 ( P = .002) 和视觉肺气肿评分 ( P < .001) 的扩散能力的相对贡献最大。

结论

归一化连接计数与慢性阻塞性肺疾病严重程度的关系可能表明，对该疾病的评估取决于低衰减体素的数量或簇的大小以及这些体素的空间排列。

©北美放射学会，2021

本文提供了在线补充材料。

另见本期 Grenier 的社论。