Correlative Factors Analysis And Prediction Model Establishment Of Benignand Malignant Solitary Pulmonary Nodules

ObjectivesThe purpose of this paper is to analyze the general clinical data,laboratory examination.imaging characteristics of solitary pulmonary nodules and the imaging characteristics of three-dimensional reconstruction of pulmonary nodules,to find out the risk factors and independent risk factors for judging the benign and malignant of solitary pulmonary nodules,to establish the malignant prediction model of solitary pulmonary nodules,and to calculate the malignant probability of solitary pulmonary nodules,so as to guide clinical medicine Management of solitary pulmonary nodules by teachers and workers.Methods1.To collect the general clinical data,laboratory examination,imaging features and three-dimensional reconstruction imaging features of 958 patients with solitary pulmonary nodule confirmed by operation or percutaneous lung biopsy in the Department of thoracic surgery and Respiratory Department of the Second Affiliated Hospital of Kunming Medical University from October 2012 to March 2020.2.Through the statistical analysis of the data collected by IBM spss22.0 software.single factor Logistic and multi factor Logistic regression analysis were used to find out the related factors and independent risk factors of the benign and malignant of solitary pulmonary nodule,and according to the results of multi factor Logistic regression analysis,a malignant prediction model of solitary pulmonary nodule was established to calculate the malignant probability of solitary pulmonary nodule.3.All cases included are divided into training data set and validation data set according to the proportion of 7:3.The training data set is used to build the model,and the validation data set is used to verify the accuracy of the model.ResultsThis study included 958 cases,506 cases of malignant nodules(52.82%),452 cases of benign nodules(47.18%);670 cases of training data set,288 cases of validation data set.Single factor Logistic regression analysis showed that:age,gender,respiratory disease history,malignant tumor history,CEA,NLR,nodule size,nodule diameter,edge characteristics,GGO component proportion,nodule internal characteristics,three-dimensional image vascular penetration sign were the relevant factors for the differentiation of benign and malignant SPN(P≤0.05).Multiple factor Logistic regression analysis showed that age,GGO composition,vascular tangle,air bronchogram,pleura traction and three-dimensional vascular penetration were independent risk factors for SPN,and NLR was independent protective factor for SPN differentiation(P≤0.05).The predictive regression equation of SPN was:P=ex/1+ex,e was natural logarithm,x=-3.14282+(0.03737 × age)+(0.51374× vacuole)+(1.87897 × vascular entanglement)+(0.90573 × air bronchogram)+(1.15055 × pleura traction)+(1.57413 × GGO content)+(1.4723 ×three-dimensional blood tube penetration sign)-(0.84182 × NLR).The ROC curve AUC of the SPN prediction model is 0.86,95%Cl is 0.832-0.888.When the intercept value t=0.234,the sensitivity of the model is 90.8%,the specificity is 68.4%,and the coincidence rate is 80.3%.The ROC curve AUC is 0.864,95%CI is 0.820-0.907,sensitivity is 77.2%,specificity is 86.3%,and coincidence rate is 81.6%.Conclusions1.To collect the general clinical data,laboratory examination,imaging features and three-dimensional reconstruction imaging features of 958 patients with solitary pulmonary nodule confirmed by operation or percutaneous lung biopsy in the Department of thoracic surgery and Respiratory Department of the Second Affiliated Hospital of Kunming Medical University from October 2012 to March 2020.2.Through the statistical analysis of the data collected by IBM spss22.0 software,single factor Logistic and multi factor Logistic regression analysis were used to find out the related factors and independent risk factors of the benign and malignant of solitary pulmonary nodule,and according to the results of multi factor Logistic regression analysis,a malignant prediction model of solitary pulmonary nodule was established to calculate the malignant probability of solitary pulmonary nodule.3.All cases included are divided into training data set and validation data set according to the proportion of 7:3.The training data set is used to build the model,and the validation data set is used to verify the accuracy of the model.