Related Experiment Video
Updated: Jul 31, 2025
04:48
Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography
Published on: November 30, 2022
2.8K
Deep learning methods for medical image fusion: A review
Tao Zhou1, QianRu Cheng1, HuiLing Lu2
1School of Computer Science and Engineering, North Minzu University, Yinchuan, 750021, China; Key Laboratory of Image and Graphics Intelligent Processing of State Ethnic Affairs Commission, North Minzu University, Yinchuan, 750021, China.
Computers in Biology and Medicine
|May 4, 2023
Summary
Deep learning image fusion methods are a computer vision hotspot. This review categorizes techniques, discusses medical applications, and outlines future directions for multi-modal medical image analysis.
Area of Science:
- Computer Vision
- Artificial Intelligence
- Medical Imaging
Background:
- Deep learning has emerged as a significant area of research in computer vision, particularly for image fusion.
- Image fusion aims to combine information from multiple images to create a single, more informative image.
Purpose of the Study:
- To systematically review deep learning-based image fusion methods.
- To categorize these methods based on their architecture and application.
- To discuss challenges and future prospects in medical image fusion.
Main Methods:
- Categorization of deep learning image fusion into End-to-End and Non-End-to-End approaches.
- Sub-categorization of Non-End-to-End methods into decision mapping and feature extraction.
- Sub-categorization of End-to-End methods based on network types: Convolutional Neural Network, Generative Adversarial Network, and Encoder-Decoder Network.
- Review of applications in the medical imaging field, including methods and datasets.
- Compilation of commonly used evaluation metrics for medical image fusion.
Main Results:
- Deep learning image fusion methods offer advantages in processing and feature extraction.
- A comprehensive taxonomy of deep learning fusion techniques is presented.
- Key applications and datasets within medical imaging are highlighted.
- A thorough overview of evaluation metrics provides a benchmark for performance assessment.
Conclusions:
- Deep learning-based image fusion holds significant promise for advancing multi-modal medical image analysis.
- Addressing challenges in datasets and fusion methodologies is crucial for future development.
- This review provides a valuable guide for researchers in the field.
