Purpose: We aimed to evaluate the performance of a deep learning system for differential diagnosis of lung cancer with conventional CT and FDG PET/CT using transfer learning (TL) and metadata. Methods: A total of 359 patients with a lung mass or nodule who underwent noncontrast chest CT and FDG PET/CT prior to treatment were enrolled retrospectively. All pulmonary lesions were classified by pathology (257 malignant, 102 benign). Deep learning classification models based on ResNet-18 were developed using the pretrained weights obtained from ImageNet data set. We propose a deep TL model for differential diagnosis of lung cancer using CT imaging data and metadata with SUVmax and lesion size derived from PET/CT. The area under the receiver operating characteristic curve (AUC) of the deep learning model was measured as a performance metric and verified by 5-fold cross-validation. Results: The performance metrics of the conventional CT model were generally better than those of the CT of PET/CT model. Introducing metadata with SUVmax and lesion size derived from PET/CT into baseline CT models improved the diagnostic performance of the CT of PET/CT model (AUC = 0.837 vs 0.762) and the conventional CT model (AUC = 0.877 vs 0.817). Conclusions: Deep TL models with CT imaging data provide good diagnostic performance for lung cancer, and the conventional CT model showed overall better performance than the CT of PET/CT model. Metadata information derived from PET/CT can improve the performance of deep learning systems.

Background/Aim: We explored the prediction of programmed cell death ligand 1 (PD-L1) expression level in non-small cell lung cancer using a machine learning approach with positron emission tomography/computed tomography (PET/CT)-based radiomics. Patients and Methods: A total of 312 patients (189 adenocarcinomas, 123 squamous cell carcinomas) who underwent F-18 fluorodeoxyglucose PET/CT were retrospectively analysed. Imaging biomarkers with 46 CT and 48 PET radiomic features were extracted from segmented tumours on PET and CT images using the LIFEx package. Radiomic features were ranked, and the top five best feature subsets were selected using the Gini index based on associations with PD-L1 expression in at least 50% of tumour cells. The areas under the receiver operating characteristic curves (AUCs) of binary classifications afforded by several machine learning algorithms (random forest, neural network, Naïve Bayes, logistic regression, adaptive boosting, stochastic gradient descent, support vector machine) were compared. The model performances were tested by 10-fold cross validation. Results: We developed and validated a PET/CT-based radiomic model predicting PD-L1 expression levels in lung cancer. Long run high grey-level emphasis, homogeneity, mean Hounsfield unit, long run emphasis from CT, and maximum standardised uptake value from PET were the five best feature subsets for positive PD-L1 expression. The Naïve Bayes model (AUC=0.712), with a sensitivity of 75.3% and specificity of 58.2%, outperformed all other classifiers. It was followed by the neural network model (AUC=0.711), random forest (AUC=0.700), logistic regression (AUC=0.673) and adaptive boosting (AUC=0.604). Conclusion: PET/CT-based radiomic features may help clinicians identify tumours with positive PD-L1 expression in a non-invasive manner using machine learning algorithms.

Purpose: Considerable discrepancies are observed between clinical staging and pathological staging after surgical resection in patients with esophageal squamous cell carcinoma (ESCC). In this study, we examined the relationships between tumor SUVs on FDG PET/CT and aggressive pathological features in resected ESCC patients. Methods: A total of 220 patients with surgically resected clinical stage I–II ESCC without neoadjuvant treatment were retrospectively analyzed. SUVmax of the primary tumor was measured on pretreatment FDG PET/CT. Pathological features included depth of tumor invasion, lymph node metastasis, tumor differentiation, lymphatic vessel tumor embolus, perineural invasion, Ki-67 index, and p53 protein expression. Receiver operating characteristic curve analysis was used to determine an optimal cutoff of SUVmax to predict pathologically advanced disease. Differences in pathological features associated with SUVmax were examined by t test or χ² test. Results: The number of patients upstaged from clinical stage I–II to pathological stage III–IV was 43 (19.5%). Receiver operating characteristic curve analysis showed that the optimal cutoff SUVmax of 4.0 had good performance for predicting locally advanced disease (area under the receiver operating characteristic curve = 0.844, P < 0.001). Higher tumor SUVmax was significantly associated with advanced depth of tumor invasion (deeper than submucosa, P < 0.001), positive lymph node metastasis (P < 0.001), presence of lymphatic vessel tumor embolus (P < 0.001), presence of perineural invasion (P < 0.001), higher Ki-67 index (P = 0.025), and poor tumor differentiation (P = 0.039). Conclusions: SUVmax measured on pretreatment FDG PET/CT is significantly associated with aggressive pathological features and may help clinicians identify patients at risk of advanced disease.

Introduction: The prognostic value of F-18 fluorodeoxyglucose positron emission tomography/computed tomography (FDG PET/CT) in hepatocellular carcinoma (HCC) was established in previous reports. However, there is no evidence suggesting the prognostic value of transcriptomes associated with tumor FDG uptake in HCC. It was aimed to elucidate metabolic genes and functions associated with FDG uptake, followed by assessment of those prognostic value.Methods: Sixty HCC patients with Edmondson–Steiner grade II were included. FDG PET/CT scans were performed before any treatment. RNA sequencing data were obtained from tumor and normal liver tissue. Associations between each metabolism-associated gene and tumor FDG uptake were investigated by Pearson correlation analyses. A novel score between glucose and lipid metabolism-associated gene expression was calculated. In The Cancer Genome Atlas Liver Hepatocellular Carcinoma dataset, the prognostic power of selected metabolism-associated genes and a novel score was evaluated for external validation.Results: Nine genes related to glycolysis and the HIF-1 signaling pathway showed positive correlations with tumor FDG uptake; 21 genes related to fatty acid metabolism and the PPAR signaling pathway demonstrated negative correlations. Seven potential biomarker genes, PFKFB4, ALDOA, EGLN3, EHHADH, GAPDH, HMGCS2, and ENO2 were identified. A metabolic gene expression balance score according to the dominance between glucose and lipid metabolism demonstrated good prognostic value in HCC.Conclusions: The transcriptomic evidence of this study strongly supports the prognostic power of FDG PET/CT and indicates the potential usefulness of FDG PET/CT imaging biomarkers to select appropriate patients for metabolism-targeted therapy in HCC.