Objectives: To use convolutional neural network for fully automated segmentation and radiomics features extraction of hypopharyngeal cancer (HPC) tumor in MRI. Methods: MR images were collected from 222 HPC patients, among them 178 patients were used for training, and another 44 patients were recruited for testing. U-Net and DeepLab V3 + architectures were used for training the models. The model performance was evaluated using the dice similarity coefficient (DSC), Jaccard index, and average surface distance. The reliability of radiomics parameters of the tumor extracted by the models was assessed using intraclass correlation coefficient (ICC). Results: The predicted tumor volumes by DeepLab V3 + model and U-Net model were highly correlated with those delineated manually (p < 0.001). The DSC of DeepLab V3 + model was significantly higher than that of U-Net model (0.77 vs 0.75, p < 0.05), particularly in those small tumor volumes of < 10 cm³ (0.74 vs 0.70, p < 0.001). For radiomics extraction of the first-order features, both models exhibited high agreement (ICC: 0.71–0.91) with manual delineation. The radiomics extracted by DeepLab V3 + model had significantly higher ICCs than those extracted by U-Net model for 7 of 19 first-order features and for 8 of 17 shape-based features (p < 0.05). Conclusion: Both DeepLab V3 + and U-Net models produced reasonable results in automated segmentation and radiomic features extraction of HPC on MR images, whereas DeepLab V3 + had a better performance than U-Net. Clinical relevance statement: The deep learning model, DeepLab V3 + , exhibited promising performance in automated tumor segmentation and radiomics extraction for hypopharyngeal cancer on MRI. This approach holds great potential for enhancing the radiotherapy workflow and facilitating prediction of treatment outcomes. Key Points: • DeepLab V3 + and U-Net models produced reasonable results in automated segmentation and radiomic features extraction of HPC on MR images. • DeepLab V3 + model was more accurate than U-Net in automated segmentation, especially on small tumors. • DeepLab V3 + exhibited higher agreement for about half of the first-order and shape-based radiomics features than U-Net.

We investigated the relationships of cystic nodal metastasis, human papillomavirus (HPV) status, and treatment failure patterns in patients with oropharyngeal squamous cell carcinoma (OPSCC) treated with chemoradiotherapy. We retrospectively reviewed pretreatment MRI and clinical courses of patients with OPSCC whose tumors were tested for HPV-induced p16 expression via immunohistochemistry and who completed chemoradiotherapy. Cervical cystic nodal metastasis and necrotic nodal metastasis were classified on MRI. Of 98 patients eligible for analysis, 33 were p16-positive. Cystic nodal metastasis was significantly more prevalent in p16-positive than in p16-negative patients (39.4% versus 18.5%, respectively; p = 0.025). Necrotic nodal metastasis was significantly more prevalent in p16-negative than in p16-positive patients (73.8% versus 51.5%, respectively; p = 0.027). On multivariate analysis, necrotic nodal metastasis (odds ratio [OR] = 7.310, p = 0.011) was an independent predictor of regional failure, while advanced nodal stage (OR = 4.119, p = 0.022) and cystic nodal metastases (OR = 0.087, p = 0.026) were independent positive and negative predictors of distant failure, respectively. Cervical cystic and necrotic nodal metastases are associated with HPV-induced p16-positive and p16-negative OPSCC, respectively. Patients with necrotic nodal metastasis at presentation have an increased risk of regional failure. Distant failure is directly and inversely correlated with advanced nodal stage and cystic nodal metastasis, respectively.

Alterations in metabolism following radiotherapy affect therapeutic efficacy, although the mechanism underlying such alterations is unclear. A new imaging technique—named dynamic nuclear polarization (DNP) carbon-13 magnetic resonance imaging (MRI)—probes the glycolytic flux in a real-time, dynamic manner. The [1-¹³C]pyruvate is transported by the monocarboxylate transporter (MCT) into cells and converted into [1-¹³C]lactate by lactate dehydrogenase (LDH). To capture the early glycolytic alterations in the irradiated cancer and immune cells, we designed a preliminary DNP ¹³C-MRI study by using hyperpolarized [1-¹³C]pyruvate to study human FaDu squamous carcinoma cells, HMC3 microglial cells, and THP-1 monocytes before and after irradiation. The pyruvate-to-lactate conversion rate (k_PL [Pyr.]) calculated by kinetic modeling was used to evaluate the metabolic alterations. Western blotting was performed to assess the expressions of LDHA, LDHB, MCT1, and MCT4 proteins. Following irradiation, the pyruvate-to-lactate conversion rates on DNP ¹³C-MRI were significantly decreased in the FaDu and the HMC3 cells but increased in the THP-1 cells. Western blot analysis confirmed the similar trends in LDHA and LDHB expression levels. In conclusion, DNP ¹³C-MRI non-invasively captured the different glycolytic alterations among cancer and immune systems in response to irradiation, implying its potential for clinical use in the future.

The aim of this study was to investigate the hyperacute and acute changes in apparent diffusion coefficient (ADC), T1, and T2 mapping in rat kidneys after severe bilateral renal ischemic-reperfusion injury (IRI). After baseline MRI, 24 Spraque-Dawley rats with renal IRI were divided equally as group 1 (post-IRI MRI at 6 hours, days 1, 3, and 7) and groups 2, 3, and 4 (post-IRI MRI at 6 hours; 6 hours and day 1; 6 hours, days 1 and 3, respectively), while six other rats without IRI (group 5) were used as sham control. ADC, T1, and T2 values of the cortex and outer and inner stripes of outer medulla (OSOM and ISOM), and immunohistochemical studies assessing monocyte chemoattractant protein-1 (MCP-1), CD68+ cells, tubular cast formation, and collagen deposition in three zones at different time points were evaluated. Significantly reduced ADCs in OSOM and ISOM are noninvasive biomarkers denoting hyperacute damages after IRI. Linear regression analysis revealed a significant inverse correlation between 6-hour/baseline ADC ratios and MCP-1 staining (P < 0.001, r² = 0.738). ADC, T1, and T2 values are useful for assessing variable IRI changes in different layers depending on underlying microstructural and histopathological changes at different time points.

To investigate the advantage of concurrent chemotherapy with postoperative radiotherapy (RT) of oral squamous cell carcinoma (OSCC) in patients with three or more minor risk factors. Minor risk factors included pT4 disease, pN1 disease, margin ≤ 4 mm, poor differentiation, perineural invasion, vessel or lymphatic invasion, and tumor invasion depth ≥ 11 mm. Surgery was the primary treatment, followed by RT or concurrent chemoradiation (CCRT). After propensity score matching, 34 patients in each treatment group were selected for comparison. The median follow-up for living patients was 86.4 months (range: 47–189 months). The 5-year overall survival of the RT and CCRT groups was 35.3% and 67.2% (p = 0.018), respectively. The 5-year recurrence-free survival of the RT group and CCRT group was 42.6% and 75.4% (p < 0.01). Postoperative CCRT for patients with three or more minor risk factors increased recurrence-free and overall survival.