Related Experiment Video
Updated: Jul 16, 2025
10:48
PET and MRI Guided Irradiation of a Glioblastoma Rat Model Using a Micro-irradiator
Published on: December 28, 2017
9.5K
Deep learning in MRI-guided radiation therapy: A systematic review.
Zach Eidex1,2, Yifu Ding1, Jing Wang1
1Department of Radiation Oncology and Winship Cancer Institute, Emory University, Atlanta, Georgia, USA.
Journal of Applied Clinical Medical Physics
|September 15, 2023
Summary
Deep learning advances in MRI-guided radiation therapy enable fully adaptive treatments. This review covers segmentation, synthesis, radiomics, and real-time MRI, highlighting challenges and trends for improved cancer care.
Area of Science:
- Medical Physics
- Radiology
- Artificial Intelligence
Background:
- Recent advances in MRI-guided radiation therapy (MRgRT) and deep learning foster fully adaptive radiation therapy (ART).
- Emerging state-of-the-art methods necessitate a systematic review of current research.
- The review covers studies published up to December 31, 2022.
Approach:
- Systematically reviewed 197 studies on deep learning in MRgRT.
- Categorized research into image segmentation, image synthesis, radiomics, and real-time MRI.
- Analyzed clinical importance and challenges of deep learning applications.
Key Points:
- Deep learning facilitates small tumor segmentation and accurate X-ray attenuation estimation from MRI.
- Radiomics aids in tumor characterization and prognosis prediction.
- Real-time MRI enables tumor motion tracking for precise treatment delivery.
- Emerging deep learning trends include multi-modal, visual transformer, and diffusion models.
Conclusions:
- Deep learning is crucial for advancing MRgRT towards fully adaptive workflows.
- Continued research is needed to address challenges in segmentation, data synthesis, and motion management.
- Future directions involve leveraging advanced deep learning architectures for enhanced cancer treatment.
