1
|
Wang H, Jin Q, Li S, Liu S, Wang M, Song Z. A comprehensive survey on deep active learning in medical image analysis. Med Image Anal 2024; 95:103201. [PMID: 38776841 DOI: 10.1016/j.media.2024.103201] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 10/20/2023] [Revised: 04/25/2024] [Accepted: 05/06/2024] [Indexed: 05/25/2024]
Abstract
Deep learning has achieved widespread success in medical image analysis, leading to an increasing demand for large-scale expert-annotated medical image datasets. Yet, the high cost of annotating medical images severely hampers the development of deep learning in this field. To reduce annotation costs, active learning aims to select the most informative samples for annotation and train high-performance models with as few labeled samples as possible. In this survey, we review the core methods of active learning, including the evaluation of informativeness and sampling strategy. For the first time, we provide a detailed summary of the integration of active learning with other label-efficient techniques, such as semi-supervised, self-supervised learning, and so on. We also summarize active learning works that are specifically tailored to medical image analysis. Additionally, we conduct a thorough comparative analysis of the performance of different AL methods in medical image analysis with experiments. In the end, we offer our perspectives on the future trends and challenges of active learning and its applications in medical image analysis. An accompanying paper list and code for the comparative analysis is available in https://github.com/LightersWang/Awesome-Active-Learning-for-Medical-Image-Analysis.
Collapse
Affiliation(s)
- Haoran Wang
- Digital Medical Research Center, School of Basic Medical Sciences, Fudan University, Shanghai 200032, China; Shanghai Key Laboratory of Medical Image Computing and Computer Assisted Intervention, Shanghai 200032, China
| | - Qiuye Jin
- Computational Bioscience Research Center (CBRC), King Abdullah University of Science and Technology (KAUST), Thuwal 23955, Saudi Arabia
| | - Shiman Li
- Digital Medical Research Center, School of Basic Medical Sciences, Fudan University, Shanghai 200032, China; Shanghai Key Laboratory of Medical Image Computing and Computer Assisted Intervention, Shanghai 200032, China
| | - Siyu Liu
- Digital Medical Research Center, School of Basic Medical Sciences, Fudan University, Shanghai 200032, China; Shanghai Key Laboratory of Medical Image Computing and Computer Assisted Intervention, Shanghai 200032, China
| | - Manning Wang
- Digital Medical Research Center, School of Basic Medical Sciences, Fudan University, Shanghai 200032, China; Shanghai Key Laboratory of Medical Image Computing and Computer Assisted Intervention, Shanghai 200032, China.
| | - Zhijian Song
- Digital Medical Research Center, School of Basic Medical Sciences, Fudan University, Shanghai 200032, China; Shanghai Key Laboratory of Medical Image Computing and Computer Assisted Intervention, Shanghai 200032, China.
| |
Collapse
|
2
|
Jin Q, Yuan M, Li S, Wang H, Wang M, Song Z. Cold-start active learning for image classification. Inf Sci (N Y) 2022. [DOI: 10.1016/j.ins.2022.10.066] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 11/05/2022]
|
3
|
Moreira da Silva DE, Gonçalves L, Franco-Gonçalo P, Colaço B, Alves-Pimenta S, Ginja M, Ferreira M, Filipe V. Active learning for data efficient semantic segmentation of canine bones in radiographs. Front Artif Intell 2022; 5:939967. [PMCID: PMC9644053 DOI: 10.3389/frai.2022.939967] [Citation(s) in RCA: 2] [Impact Index Per Article: 1.0] [Reference Citation Analysis] [Abstract] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 05/09/2022] [Accepted: 10/07/2022] [Indexed: 11/16/2022] Open
Abstract
X-ray bone semantic segmentation is one crucial task in medical imaging. Due to deep learning's emergence, it was possible to build high-precision models. However, these models require a large quantity of annotated data. Furthermore, semantic segmentation requires pixel-wise labeling, thus being a highly time-consuming task. In the case of hip joints, there is still a need for increased anatomic knowledge due to the intrinsic nature of the femur and acetabulum. Active learning aims to maximize the model's performance with the least possible amount of data. In this work, we propose and compare the use of different queries, including uncertainty and diversity-based queries. Our results show that the proposed methods permit state-of-the-art performance using only 81.02% of the data, with O(1) time complexity.
Collapse
Affiliation(s)
- D. E. Moreira da Silva
- School of Science and Technology, University of Trás-os-Montes e Alto Douro (UTAD), Vila Real, Portugal
| | - Lio Gonçalves
- School of Science and Technology, University of Trás-os-Montes e Alto Douro (UTAD), Vila Real, Portugal
- INESC Technology and Science (INESC TEC), Porto, Portugal
| | - Pedro Franco-Gonçalo
- Department of Veterinary Science, UTAD, Vila Real, Portugal
- Veterinary and Animal Research Centre (CECAV), Vila Real, Portugal
| | - Bruno Colaço
- Veterinary and Animal Research Centre (CECAV), Vila Real, Portugal
- Department of Animal Science, UTAD, Vila Real, Portugal
- Centre for the Research and Technology of Agro-Environmental and Biological Sciences (CITAB), Vila Real, Portugal
| | - Sofia Alves-Pimenta
- Veterinary and Animal Research Centre (CECAV), Vila Real, Portugal
- Department of Animal Science, UTAD, Vila Real, Portugal
- Centre for the Research and Technology of Agro-Environmental and Biological Sciences (CITAB), Vila Real, Portugal
| | - Mário Ginja
- Department of Veterinary Science, UTAD, Vila Real, Portugal
- Veterinary and Animal Research Centre (CECAV), Vila Real, Portugal
- Centre for the Research and Technology of Agro-Environmental and Biological Sciences (CITAB), Vila Real, Portugal
| | | | - Vitor Filipe
- School of Science and Technology, University of Trás-os-Montes e Alto Douro (UTAD), Vila Real, Portugal
- INESC Technology and Science (INESC TEC), Porto, Portugal
- *Correspondence: Vitor Filipe
| |
Collapse
|
4
|
CGSNet: Contrastive Graph Self-Attention Network for Session-based Recommendation. Knowl Based Syst 2022. [DOI: 10.1016/j.knosys.2022.109282] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 11/19/2022]
|
5
|
Suggestive Data Annotation for CNN-Based Building Footprint Mapping Based on Deep Active Learning and Landscape Metrics. REMOTE SENSING 2022. [DOI: 10.3390/rs14133147] [Citation(s) in RCA: 2] [Impact Index Per Article: 1.0] [Reference Citation Analysis] [Abstract] [Track Full Text] [Subscribe] [Scholar Register] [Indexed: 12/31/2022]
Abstract
Convolutional neural network (CNN)-based very high-resolution (VHR) image segmentation has become a common way of extracting building footprints. Despite publicly available building datasets and pre-trained CNN models, it is still necessary to prepare sufficient labeled image tiles to train CNN models from scratch or update the parameters of pre-trained CNN models to extract buildings accurately in real-world applications, especially the large-scale building extraction, due to differences in landscapes and data sources. Deep active learning is an effective technique for resolving this issue. This study proposes a framework integrating two state-of-the-art (SOTA) models, U-Net and DeeplabV3+, three commonly used active learning strategies, (i.e., margin sampling, entropy, and vote entropy), and landscape characterization to illustrate the performance of active learning in reducing the effort of data annotation, and then understand what kind of image tiles are more advantageous for CNN-based building extraction. The framework enables iteratively selecting the most informative image tiles from the unlabeled dataset for data annotation, training the CNN models, and analyzing the changes in model performance. It also helps us to understand the landscape features of iteratively selected image tiles via active learning by considering building as the focal class and computing the percent, the number of patches, edge density, and landscape shape index of buildings based on labeled tiles in each selection. The proposed method was evaluated on two benchmark building datasets, WHU satellite dataset II and WHU aerial dataset. Models in each iteration were trained from scratch on all labeled tiles. Experimental results based on the two datasets indicate that, for both U-Net and DeeplabV3+, the three active learning strategies can reduce the number of image tiles to be annotated and achieve good model performance with fewer labeled image tiles. Moreover, image tiles with more building patches, larger areas of buildings, longer edges of buildings, and more dispersed building distribution patterns were more effective for model training. The study not only provides a framework to reduce the data annotation efforts in CNN-based building extraction but also summarizes the preliminary suggestions for data annotation, which could facilitate and guide data annotators in real-world applications.
Collapse
|