Figure - available from: International Journal of Dentistry
This content is subject to copyright. Terms and conditions apply.
Enhanced-contrast computed tomography (CECT) showing parotid glands structures before chemoradiation therapy (CRT) in the axial slice (a). CECT showing the parotid glands delineation after the segmentation process with the aid of computer software (b).
Source publication
The multimodal approach for patients with head and neck cancer (HNC) includes treatment with chemoradiation therapy (CRT). A common concern regarding CRT side effects is the occurrence of structural and physiological alterations of the salivary glands due to exposure to ionizing radiation. The aim of this study is to examine the morphology, volume,...
Citations
... Salivary gland dysfunction after radiation therapy is a serious side effect for head and neck cancer patients that often persists years after treatment and results in diminished quality of life [56]. Currently, long-term treatments are not commonly prescribed due to adverse side effects, short-term treatments are often ineffective and only target the symptoms, and there is no standard of care for restoration of function. ...
Head and neck cancer treatment often consists of surgical resection of the tumor followed by ionizing radiation (IR), which can damage surrounding tissues and cause adverse side effects. The underlying mechanisms of radiation-induced salivary gland dysfunction are not fully understood, and treatment options are scarce and ineffective. The wound healing process is a necessary response to tissue injury, and broadly consists of inflammatory, proliferative, and redifferentiation phases with immune cells playing key roles in all three phases. In this study, select immune cells were phenotyped and quantified, and certain cytokine and chemokine concentrations were measured in mouse parotid glands after IR. Further, we used a model where glandular function is restored to assess the immune phenotype in a regenerative response. These data suggest that irradiated parotid tissue does not progress through a typical inflammatory response observed in wounds that heal. Specifically, total immune cells (CD45+) decrease at days 2 and 5 following IR, macrophages (F4/80+CD11b+) decrease at day 2 and 5 and increase at day 30, while neutrophils (Ly6G+CD11b+) significantly increase at day 30 following IR. Additionally, radiation treatment reduces CD3- cells at all time points, significantly increases CD3+/CD4+CD8+ double positive cells, and significantly reduces CD3+/CD4-CD8- double negative cells at day 30 after IR. Previous data indicate that post-IR treatment with IGF-1 restores salivary gland function at day 30, and IGF-1 injections attenuate the increase in macrophages, neutrophils, and CD4+CD8+ T cells observed at day 30 following IR. Taken together, these data indicate that parotid salivary tissue exhibits a dysregulated immune response following radiation treatment which may contribute to chronic loss of function phenotype in head and neck cancer survivors.
... Using a variety of imaging modalities such as CT, PET-CT, and MRI, several recent studies investigated smaller subject populations (ranging from n = 16 to n = 240) in different clinical and experimental settings [12][13][14][15][34][35][36]. Their details are shown in Table 4. ...
... Similar significant positive correlations between BMI and gland size were described by research groups led by Dos Santos, Heo, and Ono et al. who also suggested that fat deposits are more likely to contribute to the total volume [15,35,36]. A positive association between BMI and parotid fat saturation in MRI has been observed [47,48]. ...
Representative epidemiologic data on the average volume of the parotid gland in a large population-based MRI survey is non-existent. Within the Study of Health in Pomerania (SHIP), we examined the parotid gland in 1725 non-contrast MRI-scans in T1 weighted sequence of axial layers. Thus, a reliable standard operating procedure (Intraclass Correlation Coefficient > 0.8) could be established. In this study, we found an average, single sided parotid gland volume of 27.82 cm3 (95% confidence interval (CI) 27.15 to 28.50) in male and 21.60 cm3 (95% CI 21.16 to 22.05) in female subjects. We observed positive associations for age, body mass index (BMI), as well as male sex with parotid gland size in a multivariate model. The prevalence of incidental tumors within the parotid gland regardless of dignity was 3.94% in the Northeast German population, slightly higher than assumed. Further epidemiologic investigations regarding primary salivary gland diseases are necessary.
... For example, Matsuo et al. [22] reported that the deep learning method could discriminate benign and malignant PGTs in MRI images, with an AUC of 0.86. Gabelloni et al. [24] used magnetic resonance radiomics to discriminate PGTs, with results showing that radiomics analysis had a high diagnostic performance in pleomorphic adenomas and malignant tumors (sensitivity, specificity, and diagnostic accuracy of 0.66, 0.87, and 0.80, respectively). However, no study has explored whether deep learning can be applied in US images to differentiate benign from malignant tumors. ...
Objectives:
Evidence suggests that about 80% of all salivary gland tumors involve the parotid glands, with approximately 20% of parotid gland tumors (PGTs) being malignant. Discriminating benign and malignant parotid gland lesions preoperatively is vital for selecting the appropriate treatment strategy. This study explored the diagnostic performance of deep learning system for discriminating benign and malignant PGTs in ultrasonography images and compared it with radiologists. Methods. A total of 251 consecutive patients with surgical resection and proven parotid gland malignant or benign tumors who underwent preoperative ultrasound examinations were enrolled in this study between January 2014 and November 2020. Next, we compared the diagnostic accuracy of deep learning methods (ViT-B\16, EfficientNetB3, DenseNet121, and ResNet50) and radiologists in parotid gland tumor. In addition, the area under the curve (AUC), specificity, sensitivity, positive predictive value, and negative predictive value were calculated.
Results:
Among the 251 patients, 176/251 were the training set, whereas 75/251 were the validation set. Results showed that 74/251 patients had malignant tumor. Deep learning models achieved good performance in differentiating benign from malignant tumors, with the diagnostic accuracy and AUCs of ViT-B\16, EfficientNetB3, DenseNet121, and ResNet50 model being 81% and 0.81, 80% and 0.82, 77% and 0.81, and 79% and 0.80, respectively. On the other hand, the diagnostic accuracy and AUCs of radiologists were 77%-81% and 0.68-0.75, respectively. It was evident that the diagnostic accuracy of deep learning methods was higher than that of inexperienced radiologists, but there was no significant difference between deep learning methods and experienced radiologists.
Conclusions:
This study shows that the deep learning system can be used for diagnosing parotid tumors. The findings also suggest that the deep learning system may improve the diagnosis performance of inexperienced radiologists.
Objectives:
To evaluate the effectiveness of machine learning models based on morphological magnetic resonance imaging (MRI) radiomics in the classification of parotid tumors.
Methods:
In total, 298 patients with parotid tumors were randomly assigned to a training and test set at a ratio of 7:3. Radiomics features were extracted from the morphological MRI images and screened using the Select K Best and LASSO algorithm. Three-step machine learning models with XGBoost, SVM, and DT algorithms were developed to classify the parotid neoplasms into four subtypes. The ROC curve was used to measure the performance in each step. Diagnostic confusion matrices of these models were calculated for the test cohort and compared with those of the radiologists.
Results:
Six, twelve, and eight optimal features were selected in each step of the three-step process, respectively. XGBoost produced the highest area under the curve (AUC) for all three steps in the training cohort (0.857, 0.882, and 0.908, respectively), and for the first step in the test cohort (0.826), but produced slightly lower AUCs than SVM in the latter two steps in the test cohort (0.817 vs. 0.833, and 0.789 vs. 0.821, respectively). The total accuracies of XGBoost and SVM in the confusion matrices (70.8% and 59.6%) outperformed those of DT and the radiologist (46.1% and 49.2%).
Conclusion:
This study demonstrated that machine learning models based on morphological MRI radiomics might be an assistive tool for parotid tumor classification, especially for preliminary screening in absence of more advanced scanning sequences, such as DWI.
Key points:
• Machine learning algorithms combined with morphological MRI radiomics could be useful in the preliminary classification of parotid tumors. • XGBoost algorithm performed better than SVM and DT in subtype differentiation of parotid tumors, while DT seemed to have a poor validation performance. • Using morphological MRI only, the XGBoost and SVM algorithms outperformed radiologists in the four-type classification task for parotid tumors, thus making these models a useful assistant diagnostic tool in clinical practice.
Objective
To evaluate the influence of post-label delay times (PLDs) on the performance of 3D pseudo-continuous arterial spin labeling (pCASL) magnetic resonance imaging for characterizing parotid gland tumors and to explore the optimal PLDs for the differential diagnosis.Materials and methodFifty-eight consecutive patients with parotid gland tumors were enrolled, including 33 patients with pleomorphic adenomas (PAs), 16 patients with Warthin’s tumors (WTs), and 9 patients with malignant tumors (MTs). 3D pCASL was scanned for each patient five times, with PLDs of 1025 ms, 1525 ms, 2025 ms, 2525 ms, and 3025 ms. Tumor blood flow (TBF) was calculated, and compared among different PLDs and tumor groups. Performance of TBF at different PLDs was evaluated using receiver operating characteristic analysis.ResultsWith an increasing PLD, TBF tended to gradually increase in PAs (p < 0.001), while TBF tended to slightly increase and then gradually decrease in WTs (p = 0.001), and PAs showed significantly lower TBF than WTs at all 5 PLDs (p < 0.05). PAs showed significantly lower TBF than MTs at 4 PLDs (p < 0.05), except at 3025 ms (p = 0.062). WTs showed higher TBF than MTs at all 5 PLDs; however, differences did not reach significance (p > 0.05). Setting a TBF of 64.350 mL/100g/min at a PLD of 1525 ms, or a TBF of 23.700 mL/100g/min at a PLD of 1025 ms as the cutoff values, optimal performance could be obtained for differentiating PAs from WTs (AUC = 0.905) or from MTs (AUC = 0.872).Conclusions
Short PLDs (1025 ms or 1525 ms) are suggested to be used in 3D pCASL for characterizing parotid gland tumors in clinical practice.Key Points
• With 5 different PLDs, 3D pCASL can reflect the variation of blood flow in parotid gland tumors.
• 3D pCASL is useful for characterizing PAs from WTs or MTs.
• Short PLDs (1025 ms or 1525 ms) are suggested to be used in 3D pCASL for characterizing parotid gland tumors in clinical practice.
