基于胸部CT参数的COPD相关肺动脉高压预测模型构建OA
Construction of a predictive model for chronic obstructive pulmonary disease-related pulmonary hypertension based on chest CT parame-ters
目的 基于胸部CT参数构建慢性阻塞性肺疾病(COPD)相关肺动脉高压(COPD-PH)的列线图预测模型.方法 选择2023年1月-2024年12月就诊于青岛市市立医院的212例单纯COPD和COPD-PH病人,其中的120例作为建模集(单纯COPD 30例,COPD-PH 90例),92例作为验证集(单纯COPD 20例,COPD-PH 72例).收集病人的右下肺动脉/支气管直径比(PDRL/BDRL)、左下肺动脉/支气管直径比(PDLL/BDLL)、肺动脉/升主动脉直径比(PAD/AAD),分析其与肺动脉收缩压(PASP)的相关性.基于 CT的肺血管重构指标构建预测COPD-PH的列线图.结果 单纯COPD和不同严重程度COPD-PH病人的预测一秒用力呼气容积百分比(FEV1%)、最大呼气中期流量百分比(MMEF%)、PASP、全球慢性阻塞性肺病倡议分级差异有显著性(F=3.614~395.473,x2=27.763,P<0.05),与单纯COPD组相比,中度、重度COPD-PH组病人的FEV1%及MMEF%均显著降低(P<0.05).COPD-PH 病人 PAD/AAD、PDRL/BDRL、PDLL/BDLL与 PASP 均呈正相关(r=0.439~0.672,P<0.05).由PDRL/BDRL与PAD/AAD构建的预测COPD-PH的列线图模型在建模集和验证集中的受试者工作特征曲线下面积分别为0.894和0.842.校准曲线和决策曲线均显示列线图具有较好的预测效能.结论 基于PDRL/BDRL和PAD/AAD构建的预测COPD-PH的列线图模型具有较好的预测效能,可为临床早期筛查提供参考.
Objective To construct a nomogram prediction model for chronic obstructive pulmonary disease(COPD)-related pulmonary hypertension(COPD-PH)based on chest CT parameters.Methods A total of 212 patients with simple COPD and COPD-PH who attended Qingdao Municipal Hospital from January 2023 to December 2024 were enrolled,with 120 patients(30 with simple COPD and 90 with COPD-PH)in the modeling set and 92 patients(20 with simple COPD and 72 with COPD-PH)in the validation set.Related chest CT parameters were collected,including right lower pulmonary artery/bronchus diameter ratio(PDRL/BDRL),left lower pulmonary artery/bronchus diameter ratio(PDLL/BDLL),and pulmonary artery/ascending aortic diame-ter ratio(PAD/AAD),and their correlation with pulmonary artery systolic pressure(PASP)was analyzed.A nomogram model for predicting COPD-PH was constructed based on CT-derived pulmonary vascular remodeling indicators.Results There were sig-nificant differences between the patients with simple COPD and those with different severities of COPD-PH in predicted percentage of forced expiratory volume in 1 second(FEV1%),percentage of predicted maximum mid-expiratory flow(MMEF%),PASP,and Global Initiative for Chronic Obstructive Lung Disease staging(F=3.614-395.473,x2=27.763,P<0.05).Compared with the simple COPD group,the moderate and severe COPD-PH groups had significant reductions in FEV1%and MMEF%(P<0.05).In the patients with COPD-PH,PAD/AAD,PDRL/BDRL,and PDLL/BDLL were positively correlated with PASP(r=0.439-0.672,P<0.05).The nomogram model constructed based on PDRL/BDRL and PAD/AAD for predicting COPD-PH had an area un-der the ROC curve of 0.894 in the modeling set and 0.842 in the validation set.Both calibration curve and decision curve showed that the nomogram had good predictive performance.Conclusion The nomogram model based on PDRL/BDRL and PAD/AAD for pre-dicting COPD-PH has good predictive performance and can provide a reference for early clinical screening.
孙丽娟;刘洁;牟朝辉;韩伟;郝万明;李永春
青岛大学青岛医学院,山东青岛 266071青岛大学青岛医学院,山东青岛 266071青岛大学青岛医学院,山东青岛 266071青岛大学青岛医学院,山东青岛 266071||康复大学青岛医院(青岛市市立医院)呼吸与危重症医学科青岛大学青岛医学院,山东青岛 266071||康复大学青岛医院(青岛市市立医院)呼吸与危重症医学科青岛大学青岛医学院,山东青岛 266071||康复大学青岛医院(青岛市市立医院)呼吸与危重症医学科
医药卫生
肺疾病,慢性阻塞性高血压,肺性体层摄影术,X线计算机超声心动描记术,多普勒预测
pulmonary disease,chronic obstructivehyper-tension,pulmonarytomography,X-ray computedechocardio-graphy,Dopplerforecasting
《青岛大学学报(医学版)》 2026 (3)
428-434,7
山东省医药卫生科技重点项目(202403020-661)青岛市医药卫生科研指导项目(2024-WJKY003)
评论