摘要
目的 研究血清肺肿瘤标志物与生化、血常规指标联合检测在肺癌诊断中的价值.方法 纳入肺癌患者和肺部良性病变患者各117例,收集患者治疗前的血常规、生化和血清肺肿瘤标志物的结果.将患者分为训练集和验证集,在训练集中利用Logistic回归筛选指标建立风险预测模型,绘制ROC和Calibration曲线,并在验证集中对模型进行验证.结果 在肺部良性病变组中,ALT、AST、GGT、TBA、TBIL和DBIL之间呈正相关,而在肺癌组中,ALT、AST、GGT、ALP、TBIL、DBIL、LDH和CYFRA21-1之间呈正相关.通过Logistic回归筛选实验室指标并优选最佳预测模型T4-Model(CEA+CYFRA21-1+LYM％+NEU％).其在训练集和验证集中的AUC分别为0.808和0.782.Calibration曲线显示其在训练集和验证集上具有良好的一致性.结论 T4-Model对肺癌表现出较好的预测能力,可提高肺癌诊断效能.
Abstract
Objective To study the value of combined detection of serum lung tumor markers,biochemical and blood routine indicators in the diagnosis of lung cancer.Methods A total of 117 patients with lung cancer and 117 patients with benign pulmonary lesions were enrolled.The results of blood routine,biochemical and serum lung tumor markers were collected before treatment.The patients were divided into training set and validation set.Logistic regression was used to establish a risk prediction model in the training set,ROC curve and Calibration curve were drawn,and the model was verified in the validation set.Results Positive correlations were found among ALT,AST,GGT,TBA,TBIL and DBIL in the benign pulmonary lesion group,while positive correlations were found among ALT,AST,GGT,ALP,TBIL,DBIL,LDH and CYFRA21-1 in the lung cancer group.The laboratory indicators were screened by Logistic regression and the best prediction model T4-Model(CEA+CYFRA21-1+LYM％+NEU％)was selected.Its AUC in the training set and validation set were 0.808 and 0.782,respectively.The Calibration curve showed good agreement between the training set and the validation set.Conclusion The T4-Model has a good predictive ability for lung cancer and can improve the diagnostic efficiency of lung cancer.
基金项目
浙江省自然科学基金探索项目(LQ21H280006)
维普
万方数据