The Value Of ¹⁸F-FDG PET/CT In The Diagnosis Of Solitary Pulmonary Lesions

Objective:By analyzing the 18F-FDG PET/CT metabolic parameters MTV,SUVmax,SUVpeak,SUVmean,SUVtlg,SULmax,SULpeak,SULmean,SULtlg in solitary pulmonary lesions.?1?To analyze the diagnostic efficiency of ¹⁸F-FDG PET/CT metabolic parameters in differential diagnosis of solitary pulmonary lesions.?2?To constructe support vector machine model based on these metabolic parameters to differentiate diagnosis of solitary pulmonary lesions.Materials and Methods:We retrospectively analysed¹⁸F-FDG PET/CT scan of 135patients with solitary pulmonary lesions and subsequent pathology.All patients did not undergo any tumor-related treatment and biopsy before the examination.The original image of the patient was retrieved on the AW4.6 workstation,and using the PET VCAR software to measure the MTV,SUVmax,SUVpeak,SUVmean,SUVtlg,SULmax,SULpeak,SULmean and SULtlg,the way to delineate the ROI was 42%of maxium threshold.?1?The Mann-Whitney U nonparametric test was used to compare the differences in metabolic parameters between benign and malignant in solitary pulmonary lesions.?2?Mann-Whitney U non-parametric test was used to compare the differences between various pathological in benign groups and malignant groups.?3?The chi-square test was used to evaluate the differential diagnosis of benign and malignant solitary pulmonary lesions with SUVmax?2.5 as the criterion.The receiver operating characteristic?ROC?curve was used to evaluate the diagnostic performance of each parameter.To acquire the cutoff value when Youden index is maximum and to calculate the sensitivity,specificity,positive predictive value,negative predictive value,accuracy,and the area under the curve?AUC?of each variable is compared by the Z test.?4?The PET metabolic parameters of the lesions were obtained and used to establish the Support Vector Machine?SVM?models,which were selected according to the Akaike's information criterion?AIC?,in order to explore the value of differential diagnosis in benign and malignant solitary pulmonary lesions.The ROC curve evaluates the selected model and the permutation test was used for internal validation.Results:?1?The SUVmax,SUVpeak,SUVmean,SUVtlg,SULmax,SULpeak,SULmean,SULtlg in malignant group were higher than the benign group and the difference was statistically significant?P<0.001?,but there was no significant difference in MTV between the two groups?P=0.083?.?2?The SUVpeak,SULpeak,MTV,SUVtlg,SULtlg in inflammatory pseudotumor group were higher than the inflammatory granuloma group and the difference was statistically significant?P<0.05?.SUVtlg and SULtlg were higher in the squamous cell carcinoma group than in the adenocarcinoma group?P<0.05?.?3?The sensitivity is 100%,the specificity is only 7.41%,and the accuracy is 62.96%with SUVmax?2.5 as the criterion of benign and malignant.The cutoff value was used as the standard for the differential diagnosis of benign and malignant lesions.The AUC value?0.757-0.892?,sensitivity?70.37%-81.48%?,specificity?59.26%-90.74%?,positive Predicted values?74.44%-90.91%?,negative predictive values?62.96%-75.00%?,and accuracy?70.37%-82.22%?.?4?SVM was used to build the prediction models based on the 9metabolic parameters.Finally,two sets of optimization models were obtained?both AIC values are minimum and equal?,which were recorded as MgroupA?MTV+SUVpeak+SUVtlg?and MgroupB?MTV+SUVpeak+SULtlg?,the diagnostic accuracy was 82.96%?AUC,0.865;95%CI:0.778-0.912;sensitivity,82.72%;specificity,83.33%?,82.96%?AUC,0.863;95%CI:0.788-0.912;sensitivity,82.72%;specificity,83.33%?.Permutation test indicated both models were stable and reliable.Delong test show no difference between the two models?P=0.294?.Conclusion:¹⁸F-FDG PET/CT metabolic parameters have a good differential diagnosis for the diagnosis of solitary pulmonary lesions.Fat correction can not improve the diagnostic efficiency of metabolic parameters.The support vector machine based on the ¹⁸F-FDG PET/CT metabolic parameters can be used to predict the diagnosis of benign and malignant solitary malignant lesions and chose best parameter at the same time,which provides a new method to help differential diagnosis model and parameter screening.