Download PDFOpen PDF in browserDeep Learning for Image SegmentationEasyChair Preprint 89839 pages•Date: October 4, 2022AbstractThe purpose of this article is to explore the application value of deep learning algorithm in ultrasound images of thyroid nodules. A dataset of 7288 ultrasound images of thyroid nodules collected from the MICCAI 2020 challenge is established, the U_Net method is used, and adjustments are made on the basis of this method. Through continuous training, the optimal model is found and the computation is autonomous. accurate segmentation of thyroid nodules. The segmentation accuracy obtained by this network reaches 0.955, which has good segmentation performance. Keyphrases: Thyroid Nodules, U_Net, deep learning
|