Normal development of the velopharyngeal structures is key to obtaining good velopharyngeal closure. In the assessment of velopharyngeal closure and normal pronunciation, a variety of instruments can be used to detect and assist in the diagnosis of velopharyngeal dysfunction. In the past, the assessment of velopharyngeal closure often used two-dimensional imaging or relied solely on the subjective assessment of the phonetician. With the development of science and technology, magnetic resonance imaging (MRI) has become widely used in the evaluation of velopharyngeal structures and functions as an ideal examination method. This article reviews the current capabilities and limitations in evaluating velopharyngeal closure, as well as recent research on the structures and functions of the velopharyngeal using static MRI, dynamic MRI, three-dimensional MRI reconstructions and diffusion tensor imaging (DTI) techniques; in addition, this work explores the role and significance of MRI technology in evaluating the structures and functions of the velopharyngeal. A review of the literature shows that static MRI is simple in terms of the scanning mode, has easily adjustable parameters, and clearly shows the anatomical structures of palatopharyngeal in resting or transient vocal states. Dynamic MRI can capture the anatomical changes of the palatopharyngeal in a more complex pronunciation state and obtain accurate dynamic images of the velopharyngeal closure process for the study of speech pathology. Three-dimensional MRI reconstructions are usually used in fine scanning of the velopharyngeal structures in a resting state; although this method takes a long time, the images obtained are clear and reliable. This approach can be used for three-dimensional reconstruction analysis and three-dimensional finite element analysis, and it can be used to help plan an operation and evaluate the effect of the surgery. DTI is a new method for observing the contractile function of muscles by observing the locus of water molecules in muscles. DTI can be used to analyze and study many muscles involved in velopharyngeal closure.

Objective To evaluate the application of three deep learning algorithms in automatic segmentation of clinical target volumes (CTVs) in high-dose-rate brachytherapy after surgery for endometrial carcinoma. Methods A dataset comprising computed tomography scans from 306 post-surgery patients with endometrial carcinoma was divided into three subsets: 246 cases for training, 30 cases for validation, and 30 cases for testing. Three deep convolutional neural network models, 3D U-Net, 3D Res U-Net, and V-Net, were compared for CTV segmentation. Several commonly used quantitative metrics were employed, i.e., Dice similarity coefficient, Hausdorff distance, 95th percentile of Hausdorff distance, and Intersection over Union. Results During the testing phase, CTV segmentation with 3D U-Net, 3D Res U-Net, and V-Net showed a mean Dice similarity coefficient of 0.90 ± 0.07, 0.95 ± 0.06, and 0.95 ± 0.06, a mean Hausdorff distance of 2.51 ± 1.70, 0.96 ± 1.01, and 0.98 ± 0.95 mm, a mean 95th percentile of Hausdorff distance of 1.33 ± 1.02, 0.65 ± 0.91, and 0.40 ± 0.72 mm, and a mean Intersection over Union of 0.85 ± 0.11, 0.91 ± 0.09, and 0.92 ± 0.09, respectively. Segmentation based on V-Net was similarly to that performed by experienced radiation oncologists. The CTV segmentation time was < 3.2 s, which could save the work time of clinicians. Conclusion V-Net is better than other models in CTV segmentation as indicated by quantitative metrics and clinician assessment. Additionally, the method is highly consistent with the ground truth, reducing inter-doctor variability and treatment time.