Attentive Transfer Learning via Self-supervised Learning for Cervical Dysplasia Diagnosis
1
조회
0
다운로드
0
인용
0
참고문헌
키워드
초록
Many deep learning approaches have been studied for image classification in computer vision. However, thereare not enough data to generate accurate models in medical fields, and many datasets are not annotated. Thisstudy presents a new method that can use both unlabeled and labeled data. The proposed method is applied toclassify cervix images into normal versus cancerous, and we demonstrate the results. First, we use a patch selfsupervisedlearning for training the global context of the image using an unlabeled image dataset. Second, wegenerate a classifier model by using the transferred knowledge from self-supervised learning. We also applyattention learning to capture the local features of the image. The combined method provides better performancethan state-of-the-art approaches in accuracy and sensitivity.
Abstract
Many deep learning approaches have been studied for image classification in computer vision. However, thereare not enough data to generate accurate models in medical fields, and many datasets are not annotated. Thisstudy presents a new method that can use both unlabeled and labeled data. The proposed method is applied toclassify cervix images into normal versus cancerous, and we demonstrate the results. First, we use a patch selfsupervisedlearning for training the global context of the image using an unlabeled image dataset. Second, wegenerate a classifier model by using the transferred knowledge from self-supervised learning. We also applyattention learning to capture the local features of the image. The combined method provides better performancethan state-of-the-art approaches in accuracy and sensitivity.
참고문헌 (11)
- M. Arbyn, "Estimates of incidence and mortality of cervical cancer in 2018 : a worldwide analysis", The Lancet Global Health, 8(2), pp.e191~e203 (2020)
- T. Xu, "Multi-feature based benchmark for cervical dysplasia classification evaluation", Pattern Recognition, 63, pp.468~475 (2017)
- Z. Alyafeai, "A fully-automated deep learning pipeline for cervical cancer classification", Expert Systems with Applications, 141 (2020)[DOI]
- A. M. Vasudha, "Cervix cancer classification using colposcopy images by deep learning method", International Journal of Engineering Technology Science and Research, 5, pp.426~432 (2018)
- S. Gidaris, "Unsupervised representation learning by predicting image rotations", Proceedings of the 6th International Conference on Learning Representations (ICLR) (2018)
- T. Chen, "A simple framework for contrastive learning of visual representations", Proceedings of the 37th International Conference on Machine Learning (ICML), pp.1597~1607 (2020)
- K. He, "Momentum contrast for unsupervised visual representation learning", Proceedings of the IEEE, 2020, 9729~9738
- R. Gorantla, "Cervical cancer diagnosis using cervixnet: a deep learning approach", Proceedings of 2019 IEEE 19th International Conference on Bioinformatics and Bioengineering (BIBE), pp.397~404 (2019)
- S. Mustafa, "Evaluating convolution neural network optimization algorithms for classification of cervical cancer macro images", Proceedings of 2019 15th International Conference on Electronics, Computer and Computation (ICECCO), pp.1~5 (2019)
- V. Kudva, "Advances in Communication, Signal Processing, VLSI, and Embedded Systems", pp.299~312 (2020), Springer
- "An innovative treatment for cervical precancer (UH3)" (2017)[URL]