期刊
IEEE TRANSACTIONS ON MEDICAL IMAGING
卷 40, 期 10, 页码 2575-2588出版社
IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
DOI: 10.1109/TMI.2021.3060551
关键词
Medical image segmentation; few-shot learning; interactive learning; limited supervision
类别
资金
- National Research and Development Program of China [2019YFB1404802, 2019YFC0118802, 2018AAA0102102]
- National Natural Science Foundation of China [61672453]
- Zhejiang University Education Foundation [K18-511120-004, K17-511120017, K17-518051-02]
- Zhejiang public welfare technology research project [LGF20F020013]
- Health Research Project of Zhejiang Province of China [2019KY667]
- Wenzhou Bureau of Science and Technology of China [Y2020082]
- Key Laboratory of Medical Neurobiology of Zhejiang Province
- NSF [CCF-1617735]
This study introduces an interactive few-shot learning method named IFSL, which effectively addresses the burden of data annotation in medical image segmentation and improves model performance. Experimental results demonstrate that the IFSL method outperforms existing methods by over 20% in the DSC metric, with the interactive optimization algorithm further contributing nearly 10% improvement for few-shot segmentation models.
Many known supervised deep learning methods for medical image segmentation suffer an expensive burden of data annotation for model training. Recently, few-shot segmentation methods were proposed to alleviate this burden, but such methods often showed poor adaptability to the target tasks. By prudently introducing interactive learning into the few-shot learning strategy, we develop a novel few-shot segmentation approach called Interactive Few-shot Learning (IFSL), which not only addresses the annotation burden of medical image segmentation models but also tackles the common issues of the known few-shot segmentation methods. First, we design a new few-shot segmentation structure, called Medical Prior-based Few-shot Learning Network (MPrNet), which uses only a few annotated samples (e.g., 10 samples) as support images to guide the segmentation of query images without any pretraining. Then, we propose an Interactive Learning-based Test Time Optimization Algorithm (IL-TTOA) to strengthen our MPrNet on the fly for the target task in an interactive fashion. To our best knowledge, our IFSL approach is the first to allow few-shot segmentation models to be optimized and strengthened on the target tasks in an interactive and controllable manner. Experiments on four few-shot segmentation tasks show that our IFSL approach outperforms the state-of-the-art methods by more than 20% in the DSC metric. Specifically, the interactive optimization algorithm (IL-TTOA) further contributes similar to 10% DSC improvement for the few-shot segmentation models.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据