¹⁸F-FDG PET/CT Radiomics Nomogram For Exploring Occult Lymph Node Metastasis Of Nonsmall Cell Lung Cancer

Objective: To investigate the ability of a PET/CT-based radiomics nomogram to predict occult lymph node metastasis in patients with clinical stage N0 nonsmall cell lung cancer(NSCLC).Methods: This retrospective study collected preoperative PET and CT images and clinical data of 228 surgically confirmed non-small cell lung cancer patients scanned in our hospital from March 2012 to July 2021(training cohort,159 patients;testing cohort,69 patients).Tumor segmentation,radiomics feature extraction,and feature selection were performed on PET and CT images.radiomics models for predicting occult lymph node metastasis(OLM)was established using logistic regression algorithms based on CT,PET,and PET+CT images.Clinical predictors were selected using univariate logistic regression and multivariate logistic regression with backward selection.Finally,a nomogram was constructed by integrating the radiomics score and selected clinical predictors.Receiver operating characteristic(ROC)curves were used to verify the performance of the radiomics model and nomogram in the training and testing sets.Calibration curves assess the agreement of predicted probabilities with actual observed probabilities Decision curve is used to test whether it is clinically useful.Results: In the PET/CT based radiomics model,there was a significant difference in the radiomics score between patients with occult lymph node metastasis and those without occult lymph node metastasis(training set P<0.01,validation set P<0.01).The radiomics nomogram includes the radiomics score(Radscore)composed of six radiomics features and the location information of the tumor(central or peripheral),which demonstrated good discrimination ability in the training(area under the ROC curve [AUC] = 0.884,95% confidence interval [CI]: 0.826-0.941)and testing(AUC = 0.881,95% CI:0.8031-0.959)sets,its effect is better than three radiomics models and two metabolic parameters.Calibration curves revealed that the predicted probability of the nomogram was consistent with the actual probability of OLM.Clinical decision curve demonstrated that the nomogram was clinically useful.Conclusion: The PET/CT-based radiomics nomogram is a noninvasive tool for predicting occult lymph node metastasis in NSCLC.