Objective: To examine whether the machine-learning approach using 18-fludeoxyglucose positron emission tomography (18F-FDG-PET)-based radiomic and deep-learning features is useful for predicting the pathological risk subtypes of thymic epithelial tumors (TETs). Methods: This retrospective study included 79 TET [27 low-risk thymomas (types A, AB and B1), 31 high-risk thymomas (types B2 and B3) and 21 thymic carcinomas] patients who underwent pre-therapeutic 18F-FDG-PET/CT. High-risk TETs (high-risk thymomas and thymic carcinomas) were 52 patients. The 107 PET-based radiomic features, including SUV-related parameters [maximum SUV (SUVmax), metabolic tumor volume (MTV), and total lesion glycolysis (TLG)] and 1024 deep-learning features extracted from the convolutional neural network were used to predict the pathological risk subtypes of TETs using six different machine-learning algorithms. The area under the curves (AUCs) were calculated to compare the predictive performances. Results: SUV-related parameters yielded the following AUCs for predicting thymic carcinomas: SUVmax 0.713, MTV 0.442, and TLG 0.479 or high-risk TETs: SUVmax 0.673, MTV 0.533, and TLG 0.539. The best-performing algorithm was the logistic regression model for predicting thymic carcinomas (AUC 0.900, accuracy 81.0%), and the random forest (RF) model for high-risk TETs (AUC 0.744, accuracy 72.2%). The AUC was significantly higher in the logistic regression model than three SUV-related parameters for predicting thymic carcinomas, and in the RF model than MTV and TLG for predicting high-risk TETs (each; p < 0.05). Conclusion: 18F-FDG-PET-based radiomic analysis using a machine-learning approach may be useful for predicting the pathological risk subtypes of TETs. Advances in knowledge: Machine-learning approach using 18F-FDG-PET-based radiomic features has the potential to predict the pathological risk subtypes of TETs.

Objective: To identify risk factors for symptomatic radiation pneumonitis (RP) after stereotactic radiation therapy (SRT) for lung tumours. Methods: We retrospectively evaluated 68 lung tumours in 63 patients treated with SRT between 2011 and 2015. RP was graded according to the National Cancer Institute—Common Terminology Criteria for Adverse Events version 4.0. SRT was delivered at 7.0–12.0 Gy per each fraction, once daily, to a total of 48–64 Gy (median, 50 Gy). Univariate analysis was performed to assess patient- and treatment-related factors, including age, sex, smoking index (SI), pulmonary function, tumour location, serum Krebs von den Lungen-6 value (KL-6), dose-volume metrics (V5, V10, V20, V30, V40 and VS5), homogeneity index of the planning target volume (PTV), PTV dose, mean lung dose (MLD), contralateral MLD and V2, PTV volume, lung volume and the PTV/lung volume ratio (PTV/Lung). Performance of PTV/Lung in predicting symptomatic RP was also analysed using receiver operating characteristic (ROC) analysis. Results: The median follow-up period was 21 months. 10 of 63 patients (15.9%) developed symptomatic RP after SRT. On univariate analysis, V10, V20, PTV volume and PTV/Lung were significantly associated with occurrence of RP ≥Grade 2. ROC curves indicated that symptomatic RP could be predicted using PTV/Lung [area under curve (AUC): 0.88, confidence interval (CI: 0.78–0.95), cut-off value: 1.09, sensitivity: 90.0% and specificity: 72.4%]. Conclusion: PTV/Lung is a good predictor of symptomatic RP after SRT. Advances in knowledge: The cases with high PTV/Lung should be carefully monitored with caution for the occurrence of RP after SRT.

Prediction of tumor consistency is valuable for planning transsphenoidal surgery for pituitary adenoma. A prospective study was conducted involving 49 participants with pituitary adenoma to determine whether quantitative pharmacokinetic analysis of dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) is useful for predicting consistency of adenomas. Pharmacokinetic parameters in the adenomas including volume of extravascular extracellular space (EES) per unit volume of tissue (v_e), blood plasma volume per unit volume of tissue (v_p), volume transfer constant between blood plasma and EES (K^trans), and rate constant between EES and blood plasma (k_ep) were obtained. The pharmacokinetic parameters and the histologic percentage of collagen content (PCC) were compared between soft and hard adenomas using Mann–Whitney U test. Pearson’s correlation coefficient was used to correlate pharmacokinetic parameters with PCC. Hard adenomas showed significantly higher PCC (44.08 ± 15.14% vs. 6.62 ± 3.47%, p < 0.01), v_e (0.332 ± 0.124% vs. 0.221 ± 0.104%, p < 0.01), and K^trans (0.775 ± 0.401/min vs. 0.601 ± 0.612/min, p = 0.02) than soft adenomas. Moreover, a significant positive correlation was found between v_e and PCC (r = 0.601, p < 0.01). The v_e derived using DCE-MRI may have predictive value for consistency of pituitary adenoma.