Deep learning‐based ultrasonic dynamic video detection and segmentation of thyroid gland and its surrounding cervical soft tissues. Issue 1 (29th November 2021)

• Record Type: Journal Article
• Authors: Luo, Hongxia; Ma, Laifa; Wu, Xiangqiong; Tan, Guanghua; Zhu, Hui; Wu, Senmin; Li, Kenli; Yang, Yan; Li, Shengli
• Is Part Of: Medical physics. Volume 49:Issue 1(2022)
• Page Start: 382
• Page End: 392
• Publication Date: 2021-11-29
• Subjects: deep learning -- detection -- segmentation -- thyroid gland -- ultrasonic videos; Medical physics -- Periodicals; Medical physics; Geneeskunde; Natuurkunde; Toepassingen; Biophysics; Periodicals; Periodicals; Electronic journals; 610.153
• DOI: 10.1002/mp.15332
• Languages: English
• ISSNs: 0094-2405

Abstract

Abstract: Background: The prevalence of thyroid diseases has been increasing year by year. In this study, we established and validated a deep learning method (Cascade region‐based convolutional neural network, R‐CNN) based on ultrasound videos for automatic detection and segmentation of the thyroid gland and its surrounding tissues in order to reduce the workload of radiologists and improve the detection and diagnosis rate of thyroid disease. Methods: Seventy‐one patients with normal thyroid ultrasound were included. The ultrasound videos of 59 patients were used as the training dataset, the data of 12 patients were used as the validation dataset, and in addition, the data of 9 patents were used as the testing dataset. Ultrasound videos of thyroid examination, including five standard sections (left and right lobe transverse scan, central isthmus transverse scan, left and right lobe longitudinal scan), were collected from all patients. The radiologists labeled the neck tissues, including anterior cervical muscle, cricoid cartilage, trachea, thyroid gland, endothyroid vessels, carotid artery, internal jugular vein, and esophagus. A large dataset was constructed to train and test the deep learning method. The performance was evaluated using the COCO metrics AP, AP50, and AP75. We compared the Cascade R‐CNN with a state‐of‐the‐art method CenterMask in the test dataset. Results: We annotated 166817, 34364, and 29227 regions in training, validation and testing samples. The model could achieve a good detection performance for the thyroid left lobe, right lobe, isthmus, muscles, trachea, carotid artery, and jugular vein; the AP50 of these tissues were 86.5%, 87.5%, 89.1%, 96.1%, 96.6%, 97.7%, and 91.8%, respectively. In addition, the model showed good segmentation performance for the muscles, trachea, and carotid artery; the AP50 of these tissues were 96%, 96.6%, and 97.8%, respectively. For the left lobe, right lobe, isthmus, esophagus, and jugular vein, AP50 was ≥86%. However, the segmentation results for the cricoid cartilage and endothyroid vessels were not high (AP50 of 53.9% and 48.5%, respectively). For fair comparison, the performance of Cascade R‐CNN is better than that of CenterMask for detection and segmentation tasks. The difference was statistically significant ( p < 0.05). Conclusions: The new method could successfully detect and segment the thyroid gland and its surrounding tissues.