胡艳军,魏璇,赵鹏飞,等.低剂量胸部CT辐射风险的倾向性评分匹配研究[J].中华放射医学与防护杂志,2026,46(1):77-83.Hu Yanjun,Wei Xuan,Zhao Pengfei,et al.Propensity score matching-based study on radiation risk of low-dose chest CT[J].Chin J Radiol Med Prot,2026,46(1):77-83
低剂量胸部CT辐射风险的倾向性评分匹配研究
Propensity score matching-based study on radiation risk of low-dose chest CT
投稿时间:2025-04-18  
DOI:10.3760/cma.j.cn112271-20250418-00145
中文关键词:  倾向性评分匹配  低剂量CT  胸部CT  辐射剂量  终生归因风险  继发性肿瘤
英文关键词:Propensity score matching  Low dose CT  Chest CT  Radiation dose  Lifetime attributable risk  Secondary malignancies
基金项目:国家自然科学基金(82271988)
作者单位E-mail
胡艳军 首都医科大学附属北京友谊医院放射科, 北京 100050  
魏璇 首都医科大学附属北京友谊医院放射科, 北京 100050  
赵鹏飞 首都医科大学附属北京友谊医院放射科, 北京 100050  
苏垠平 中国疾病预防控制中心辐射防护与核安全医学所 辐射防护与核应急中国疾病预防控制中心重点实验室, 北京 100088  
邱睿 清华大学工程物理系 粒子技术与辐射成像教育部重点实验室, 北京 100084  
田心 首都医科大学附属北京友谊医院放射科, 北京 100050 ytniu163vip@163.com 
王丽坤 首都医科大学附属北京友谊医院放射科, 北京 100050  
牛延涛 首都医科大学附属北京友谊医院放射科, 北京 100050  
摘要点击次数: 4751
全文下载次数: 590
中文摘要:
      目的 使用倾向性评分匹配法比较胸部低剂量CT(LDCT)与常规剂量CT的辐射剂量及其致癌风险。方法 收集首都医科大学附属北京友谊医院放射科2022年1月至2023年12月期间,18~80岁受检者在同一设备接受胸部CT平扫检查的相关信息。先按性别将数据分为男女两个子集,再按扫描方法分为LDCT组和常规CT组,分别对男女两个子集应用倾向性评分匹配(协变量:检查年龄、受检者直径、扫描长度),获得均衡匹配的两组。采用Mann-Whitney U检验比较两组受检者分别在以下指标的差异:剂量长度积(DLP)、有效剂量(E)、肺器官剂量、乳腺器官剂量(女)、肺癌及乳腺癌终生归因风险(LAR)。结果 研究纳入男性受试者25 634例,女性受试者36 463例。匹配后,男性成功匹配10 082对,女性14 124对。与常规CT组相比,男性受检者LDCT的DLP、E、肺器官剂量、肺癌发病率LAR和肺癌死亡率LAR的差异均有统计学意义(Z=-120.6、-122.4、-122.3、-118.6、-119.8,P<0.01)。女性受检者除上述指标(Z=-140.7、-144.1、-144.3、-142.0、-142.7,P<0.01),乳腺器官剂量、乳腺癌发病率LAR和乳腺癌死亡率LAR的差异也有统计学意义(Z=-142.4、-86.0、-97.6,P<0.01)。在DLP、E、肺器官吸收剂量、肺癌LAR、乳腺器官吸收剂量和乳腺癌LAR等指标中,常规CT组是LDCT组的3.0~4.0倍。结论 LDCT降低辐射风险作用显著。性别因素显著影响辐射风险相关指标,对制定差异化诊断参考水平有借鉴意义。
英文摘要:
      Objective To compare radiation doses and carcinogenic risks between low-dose chest CT (LDCT) and standard chest CT (STD) using propensity score matching. Methods Data on chest CT scans (non-contrast-enhanced) were retrospectively collected from January 2022 to December 2023 using a standardized protocol on a single scanner. The dataset was first stratified by gender into male and female subgroups; patients were then categorized into LDCT and STD groups based on scanning protocols; propensity score matching (PSM) was subsequently applied to each gender subgroup using covariates (patient age, patient diameter, scan length), yielding balanced LDCT and STD cohorts for comparative analysis. The Mann-Whitney U test were conducted to evaluate differences in radiation metrics (DLP, effective dose E, lung organ dose, breast organ dose for females) and lifetime attributable risks (LAR) for lung cancer and breast cancer between the two groups. Results The study included 25 634 male patients and 36 463 female patients. After PSM, 10 082 matched pairs were formed in the male cohort, and 14 124 matched pairs were formed in the female cohort. In males, the differences in DLP, E, lung organ dose, lung cancer incidence LAR, and lung cancer mortality LAR between the STD and LDCT groups were statistically significant (Z = -120.6, -122.4, -122.3, -118.6, -119.8, P<0.01).In females, in addition to the aforementioned indicators (Z = -140.7, -144.1, -144.3, -142.0, -142.7, P<0.01), the differences in breast organ dose, breast cancer incidence LAR, and breast cancer mortality LAR between the STD and LDCT groups were also statistically significant (Z = -142.4, -86.0, -97.6, P<0.01). For indicators including DLP, effective dose, lung organ dose, lung cancer LAR, breast organ dose, and breast cancer LAR, the STD group was 3.0–4.0 times that of the LDCT group. Conclusions LDCT demonstrated a significant protective effect against radiation risks. Gender factors significantly influenced radiation risk-related indicators, providing critical insights for establishing gender-specific diagnostic reference levels.
HTML  查看全文  查看/发表评论  下载PDF阅读器
关闭