Identifying Malignant Breast Ultrasound Images Using ViT-Patch
-
Published:2023-03-09
Issue:6
Volume:13
Page:3489
-
ISSN:2076-3417
-
Container-title:Applied Sciences
-
language:en
-
Short-container-title:Applied Sciences
Author:
Feng Hao1, Yang Bo1ORCID, Wang Jingwen1, Liu Mingzhe2ORCID, Yin Lirong3, Zheng Wenfeng1ORCID, Yin Zhengtong4, Liu Chao5ORCID
Affiliation:
1. School of Automation Engineering, University of Electronic Science and Technology, Chengdu 610000, China 2. School of Data Science and Artificial Intelligence, Wenzhou University of Technology, Wenzhou 325000, China 3. Department of Geography and Anthropology, Louisiana State University, Baton Rouge, LA 70803, USA 4. College of Resource and Environment Engineering, Guizhou University, Guiyang 550025, China 5. LIRMM, UMR 5506, CNRS-UM, 34095 Montpellier, France
Abstract
Recently, the Vision Transformer (ViT) model has been used for various computer vision tasks, due to its advantages to extracting long-range features. To better integrate the long-range features useful for classification, the standard ViT adds a class token, in addition to patch tokens. Despite state-of-the-art results on some traditional vision tasks, the ViT model typically requires large datasets for supervised training, and thus, it still face challenges in areas where it is difficult to build large datasets, such as medical image analysis. In the ViT model, only the output corresponding to the class token is fed to a Multi-Layer Perceptron (MLP) head for classification, and the outputs corresponding to the patch tokens are exposed. In this paper, we propose an improved ViT architecture (called ViT-Patch), which adds a shared MLP head to the output of each patch token to balance the feature learning on the class and patch tokens. In addition to the primary task, which uses the output of the class token to discriminate whether the image is malignant, a secondary task is introduced, which uses the output of each patch token to determine whether the patch overlaps with the tumor area. More interestingly, due to the correlation between the primary and secondary tasks, the supervisory information added to the patch tokens help with improving the performance of the primary task on the class token. The introduction of secondary supervision information also improves the attention interaction among the class and patch tokens. And by this way, ViT reduces the demand on dataset size. The proposed ViT-Patch is validated on a publicly available dataset, and the experimental results show its effectiveness for both malignant identification and tumor localization.
Funder
Sichuan Science and Technology Support Program
Subject
Fluid Flow and Transfer Processes,Computer Science Applications,Process Chemistry and Technology,General Engineering,Instrumentation,General Materials Science
Reference33 articles.
1. Zheng, W., Yang, B., Xiao, Y., Tian, J., Liu, S., and Yin, L. (2022). Low-Dose CT Image Post-Processing Based on Learn-Type Sparse Transform. Sensors, 22. 2. Quantitative Evaluation of an Automated Cone-Based Breast Ultrasound Scanner for MRI–3D US Image Fusion;Nikolaev;IEEE Trans. Med. Imaging,2021 3. Sparse Angle CBCT Reconstruction Based on Guided Image Filtering;Xu;Front. Oncol.,2022 4. Brosch, T., and Tam, R. (2013, January 22–26). Manifold Learning of Brain MRIs by Deep Learning. Proceedings of the Medical Image Computing and Computer-Assisted Intervention (MICCAI), Nagoya, Japan. 5. Plis, S.M., Hjelm, D.R., Salakhutdinov, R., and Calhoun, V.D. (2013). Deep learning for neuroimaging: A validation study. arXiv.
Cited by
27 articles.
订阅此论文施引文献
订阅此论文施引文献,注册后可以免费订阅5篇论文的施引文献,订阅后可以查看论文全部施引文献
|
|