摘要
目的构建与应用肺小结节恶性概率预测模型,为有创诊疗提供客观依据,以避免不必要的侵袭操作及可能引发的严重后果。方法回顾性分析2019年1月至2021年3月在普宁市人民医院手术切除及经病理诊断的123例肺小结节患者的临床资料。根据病理结果将其分为良性病变组与恶性病变组,比较两组患者临床资料及影像学资料,对筛选出的恶性肺小结节预测因子进行分析,并使用二分类Logistic回归方程式构建其恶性概率预测模型,通过对该模型的应用验证其预测价值。结果 123例肺小结节患者的活检病理结果显示,良性肺小节40例(32.52%),恶性肺小节83例(67.48%)。两组性别、血管征、毛刺、边缘光滑、胸膜牵拉、分叶及肺结节最大径比较,差异有统计学意义(P <0.05)。构建出肺小结节恶性概率预测模型为P=1∕(1+e-X)。根据建立的肺小结节恶性概率预测模型选取121例受试,预测灵敏度为96%,特异性为55%,阳性预测81%,阴性预测值为86%。结论构建与应用肺小结节恶性概率预测模型价值明显,可作为手术切除前的初筛方法。
Objective To construct and apply the malignant probability prediction model of pulmonary nodules,provide objective basis for invasive diagnosis and treatment,and avoid unnecessary invasive operations and possible serious consequences. Methods The clinical data of 123 patients with pulmonary nodules who underwent surgical resection and pathological diagnosis in Puning People’s Hospital from January 2019 to March 2021 were analyzed retrospectively according to the pathological diagnosis,they were divided into benign lesion group and malignant lesion group. The clinical and imaging data of the two groups were compared,the predictors of malignant pulmonary nodules were analyzed,and the binary logistic regression equation was used to construct the malignant probability prediction model. The prediction value was verified by the application of the model. Results After biopsy pathological examination of 123 patients with pulmonary nodules,40 patients( 32. 52%) with benign pulmonary nodules and 83 patients( 67. 48%) with malignant pulmonary nodules were found.There were significant differences in gender,vascular signs,burrs,smooth edges,pleural traction,lobulation and the maximum diameter of pulmonary nodules between the two groups( P < 0. 05). The malignant probability prediction model of pulmonary nodules was established as P = 1/( 1 + E-x). A total of 121 subjects were selected according to the established malignant probability prediction model of pulmonary nodules. The prediction sensitivity was 96%,the specificity was 55%,the positive prediction was 81%,and the negative prediction was 86%. Conclusions The construction and application of the malignant probability prediction model of pulmonary nodules has obvious value,which can be used as a preliminary screening method before surgical resection.
作者
陈子伟
Chen Ziwei(Department of Respiratory Medicine,Puning People’s Hospital,Puning 515300,China)
出处
《临床医学》
CAS
2022年第1期14-16,共3页
Clinical Medicine
关键词
肺小结节
恶性概率
预测模型
应用
Pulmonary nodules
Malignant probability
Prediction model
Application
