Study On The Accuracy Of Diagnosis Of Benign And Malignant Pulmonary Nodules By CT-guided Percutaneous Transthoracic Needle Biopsy And The Influencing Factors Of Complications

Objective: To analyze the accuracy of diagnosis of pulmonary nodules by CT-guided percutaneous puncture biopsy and the influencing factors of related complications by collecting the clinical and surgical data of patients undergoing CT-guided percutaneous puncture biopsy.The model was established to identify the possibility of complications according to the influencing factors,so as to provide a basis for reducing or avoiding the occurrence of such technical complications.Methods:(1)(1)Clinical,imaging and intraoperative information of 330 outpatients and inpatients who underwent percutaneous pulmonary nodule biopsy in the Affiliated Hospital of Qingdao University from May 2021 to February 2022 were collected and followed up 3,6 and 12 months after surgery.287 patients were finally included.Pathological results obtained by puncture were compared with postoperative pathological results or follow-up clinical diagnosis results,and descriptive statistical analysis of the data was conducted by SPSS 26.0 statistical software.The measurement data of normal distribution were expressed as Mean±SD,the T test was used for comparison between groups,the frequency/percentage was used for description of count data,and the Chi-square test was used for comparison between groups.The benign and malignant rates after biopsy were compared with those after surgery using Chi-square test or exact probability method.(2)According to the ratio of 8:2,the collected meaningful cases were divided into the training set(227 cases)and the verification set(60 cases).Univariate analysis was used to analyze the influencing factors of complications,and multivariate logistics regression was used to judge the independent predictors of the risk of common complications.P < 0.05 was considered as the statistical significance of difference.Fisher discriminant function was constructed by including statistically significant risk factors of complications,and the accuracy of the model was calculated by cross-verification method.The validation set was used for model practice and evaluation to evaluate the accuracy of the model.Results:(1)The sensitivity,specificity and accuracy of CT-guided percutaneous pulmonary puncture biopsy for the diagnosis of benign and malignant pulmonary nodules were 96.3%,100.0% and 97.6%,which had no statistical significance compared with the final clinical diagnosis.In addition,there were no significant differences in sensitivity,specificity and accuracy among different groups according to the size and nature of the lesions(P > 0.05).(2)Common complications of puncture included pneumothorax(39/227,17.2%)and pulmonary hemorrhage(100/227,44.1%),partial cough and hemoptysis(26/227,11.5%),and chest hemorrhage(2/227,0.9%).Independent risk factors for pneumothorax included emphysema and crossing the pleura more than twice.Independent risk factors for pulmonary hemorrhage included lesion size,puncture needle Angle through pleura < 50°,lesion distance from diaphragmatic surface < 3cm,and halo sign at lesion edge.The influence factors of cough and hemoptysis include distance through lung tissue and lesion location.(3)Select independent risk factors of complications: distance between cleft leaves < 3 cm,the lesion margin with halo sign,the lesions appear internal cavity,or hole,traversing lung pleura number 2 or more times,distance,build Fisher discriminant function,the accuracy of the prediction model of cross validation group was 67.0%.The accuracy was 63.3% when the case group of validation set was substituted into the prediction model.Conclusions:(1)CT guided lung biopsy in the diagnosis of pulmonary nodules benign and malignancy with high accuracy.(2)The independent risk factors of pneumothorax were: combined with emphysema,crossing the pleura ≥2 times.Lesion size,puncture needle through pleura Angle < 50°,lesion distance from diaphragmatic surface < 3cm,lesion edge halo sign were independent risk factors for pulmonary hemorrhage.The distance through lung tissue and the location of the lesion were the influencing factors of coughing and hemoptysis.(3)The function model constructed by Fisher discrimination has high accuracy in assisting to distinguish the possibility of complications.