From the UAB Lung Imaging Lab (P.R.A.P., V.L.S., A.N., A.K.P., S.P.B., S.B.), Department of Computer Science (P.R.A.P., C.Z.), Department of Electrical and Computer Engineering (V.L.S., A.N.), and Division of Pulmonary, Allergy and Critical Care Medicine (A.K.P., S.P.B., S.B.), University of Alabama at Birmingham, 1720 2nd Ave S, THT 422, Birmingham, AL 35294; and The Roy J. Carver Department of Biomedical Engineering, University of Iowa, Iowa City, Iowa (M.F.A.C., J.M.R.).
Radiology. 2023 Jun;307(5):e222998. doi: 10.1148/radiol.222998.
Background Approximately half of adults with chronic obstructive pulmonary disease (COPD) remain undiagnosed. Chest CT scans are frequently acquired in clinical practice and present an opportunity to detect COPD. Purpose To assess the performance of radiomics features in COPD diagnosis using standard-dose and low-dose CT models. Materials and Methods This secondary analysis included participants enrolled in the Genetic Epidemiology of COPD, or COPDGene, study at baseline (visit 1) and 10 years after baseline (visit 3). COPD was defined by a forced expiratory volume in the 1st second of expiration to forced vital capacity ratio less than 0.70 at spirometry. The performance of demographics, CT emphysema percentage, radiomics features, and a combined feature set derived from inspiratory CT alone was evaluated. CatBoost (Yandex), a gradient boosting algorithm, was used to perform two classification experiments to detect COPD; the two models were trained and tested on standard-dose CT data from visit 1 (model I) and low-dose CT data from visit 3 (model II). Classification performance of the models was evaluated using area under the receiver operating characteristic curve (AUC) and precision-recall curve analysis. Results A total of 8878 participants (mean age, 57 years ± 9 [SD]; 4180 female, 4698 male) were evaluated. Radiomics features in model I achieved an AUC of 0.90 (95% CI: 0.88, 0.91) in the standard-dose CT test cohort versus demographics (AUC, 0.73; 95% CI: 0.71, 0.76; < .001), emphysema percentage (AUC, 0.82; 95% CI 0.80, 0.84; < .001), and combined features (AUC, 0.90; 95% CI: 0.89, 0.92; = .16). Model II, trained on low-dose CT scans, achieved an AUC of 0.87 (95% CI: 0.83, 0.91) on the 20% held-out test set for radiomics features compared with demographics (AUC, 0.70; 95% CI: 0.64, 0.75; = .001), emphysema percentage (AUC, 0.74; 95% CI: 0.69, 0.79; = .002), and combined features (AUC, 0.88; 95% CI: 0.85, 0.92; = .32). Density and texture features were the majority of the top 10 features in the standard-dose model, whereas shape features of lungs and airways were significant contributors in the low-dose CT model. Conclusion A combination of features representing parenchymal texture and lung and airway shape on inspiratory CT scans can be used to accurately detect COPD. ClinicalTrials.gov registration no. NCT00608764 © RSNA, 2023 See also the editorial by Vliegenthart in this issue.
大约一半的慢性阻塞性肺疾病（COPD）成人患者未被诊断。在临床实践中，经常获取胸部 CT 扫描，并提供了检测 COPD 的机会。目的：使用标准剂量和低剂量 CT 模型评估放射组学特征在 COPD 诊断中的表现。
本二次分析纳入了在基线（第 1 次就诊）和基线后 10 年（第 3 次就诊）参加 COPD 遗传流行病学或 COPDGene 研究的参与者。在肺活量测定时，第 1 秒用力呼气量与用力肺活量的比值小于 0.70 定义为 COPD。评估了人口统计学数据、CT 肺气肿百分比、放射组学特征以及单独吸气 CT 衍生的组合特征集的表现。使用梯度提升算法 CatBoost（Yandex）执行了两个分类实验来检测 COPD；这两个模型分别在第 1 次就诊的标准剂量 CT 数据（模型 I）和第 3 次就诊的低剂量 CT 数据（模型 II）上进行了训练和测试。使用受试者工作特征曲线下面积（AUC）和精度-召回曲线分析评估模型的分类性能。
共评估了 8878 名参与者（平均年龄 57 岁±9[标准差]；4180 名女性，4698 名男性）。放射组学特征在标准剂量 CT 检验队列中，模型 I 的 AUC 为 0.90（95%CI：0.88，0.91），优于人口统计学（AUC：0.73；95%CI：0.71，0.76；<.001）、肺气肿百分比（AUC：0.82；95%CI 0.80，0.84；<.001）和组合特征（AUC：0.90；95%CI：0.89，0.92；=.16）。基于低剂量 CT 扫描训练的模型 II，在放射组学特征的 20%保留测试集中，对低剂量 CT 扫描的 AUC 为 0.87（95%CI：0.83，0.91），优于人口统计学（AUC：0.70；95%CI：0.64，0.75；<.001）、肺气肿百分比（AUC：0.74；95%CI：0.69，0.79；<.002）和组合特征（AUC：0.88；95%CI：0.85，0.92；=.32）。标准剂量模型中，前 10 个特征中大多数是密度和纹理特征，而低剂量 CT 模型中，肺部和气道的形状特征是重要的贡献者。
吸气 CT 扫描上代表实质纹理和肺及气道形状的特征组合可用于准确检测 COPD。
ClinicalTrials.gov 注册号：NCT00608764 © RSNA，2023 参见本期 Vliegenthart 的社论。
