Related Experiment Video
Updated: Jun 25, 2025
04:48
Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography
Published on: November 30, 2022
2.7K
A comprehensive survey on deep active learning in medical image analysis.
Haoran Wang1, Qiuye Jin2, Shiman Li1
1Digital 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.
Medical Image Analysis
|May 22, 2024
Summary
Active learning reduces medical image annotation costs by selecting informative samples for deep learning models. This survey reviews core methods, integrates them with other label-efficient techniques, and analyzes their performance in medical imaging.
Area of Science:
- Medical Image Analysis
- Artificial Intelligence
- Machine Learning
Background:
- Deep learning models require large expert-annotated datasets for medical image analysis.
- High annotation costs for medical images hinder deep learning development.
- Active learning (AL) offers a solution by minimizing required labeled data.
Purpose of the Study:
- To provide a comprehensive review of active learning methods for medical image analysis.
- To summarize the integration of active learning with other label-efficient techniques.
- To offer insights into future trends and challenges in this domain.
Main Methods:
- Review of core active learning techniques, focusing on informativeness evaluation and sampling strategies.
- Detailed summary of active learning's synergy with semi-supervised and self-supervised learning.
- Experimental comparative analysis of various active learning methods in medical image analysis.
Main Results:
- Active learning significantly reduces the need for extensive manual annotation in medical imaging.
- Integration with other label-efficient methods enhances AL's effectiveness.
- Comparative analysis provides empirical evidence for the performance of different AL strategies.
Conclusions:
- Active learning is crucial for cost-effective development of deep learning in medical image analysis.
- Future research should focus on novel AL strategies and their integration with emerging AI techniques.
- Addressing challenges in AL implementation will accelerate AI adoption in healthcare.
