Advanced Surface-Guided Radiation Therapy
Transforming surface guided radiation therapy with low-cost RGB cameras and neural networks for real-time patient tracking and internal anatomy estimation.
Recent Accomplishments
View All NewsBest in Physics award at AAPM 2023
Awarded to the top 15 abstracts out of 2200+ submissions, placing in the top 1% of all research presented at the conference.
Atharva Peshkar and Mohamed Eldib win awards for AAPM 2024
Both researchers recognized for outstanding contributions in medical physics, with awards for innovation and excellence.
Computer vision project funded by CU Anschutz Cancer Center
Secured funding for computer vision surface imaging breast DIBH project with very positive reviews from the Cancer Center.
First prize in AAPM Rocky Mountain Chapter 'Med Phys Slam'
Won first prize in the competition and represented the Rocky Mountain Chapter at the AAPM annual meeting in Houston.
Benchmarking trial begins for Computer Vision technique
Started benchmarking our Computer Vision patient alignment technique against the current gold standard IR-marker motion tracking.
Project Overview
Traditional surface-guided radiation therapy (SGRT) relies on fixed, expensive hardware with limited adaptability. Our approach replaces this with low-cost RGB cameras and neural networks, enabling real-time patient pose tracking and internal anatomy estimation from external surfaces.
Improved Access
Improves access to simulation-free treatment for ~50% of patients currently excluded
Reduced Clinic Visits
Leverages diagnostic CTs and video-based pose tracking to reduce clinic visits
Higher Precision
Models skeletal and soft tissue shifts using deformation models for higher precision
This innovation supports simulation-free workflows — eliminating the need for CT simulation in palliative treatment and reducing delays, burden, and cost.
This work builds on advanced computer vision techniques including SMPL-based modeling and pose estimation, applied in the context of real-world radiation oncology workflows.
Multi-View Demonstration
Simultaneous visualization of RGB, depth, pose tracking, and internal anatomy estimation
3D Point Cloud Visualization
Interactive 3D visualization for radiation therapy planning and analysis.
3D Viewer Preview
The 3D viewer is only available in the deployed environment.
To view the 3D model:
- Deploy this project to Vercel
- The model will load from: https://gruc9opbjll8ofcl.public.blob.vercel-storage.com/scene.glb
- Visit the deployed site to see your 3D model
Stay Updated
Subscribe to our newsletter to receive the latest updates on our research, publications, and events. We regularly share insights on radiation therapy innovations, computer vision applications, and more.
Stay Updated
Subscribe to our newsletter for the latest research updates, publications, and events.
Latest Publications
Upcoming Events
- AAPM Annual MeetingJuly 14-18, 2024
- Research SymposiumSeptember 5, 2024
- Workshop: AI in Radiation TherapyOctober 12, 2024
About the Team
This research is led by Dr. David Thomas at the Thomas Lab, Jefferson Radiation Oncology. Our team includes experts in medical physics, computer science, and biomedical engineering.
Acknowledgments
We gratefully acknowledge the contributions of the broader open-source and academic communities whose ideas, design patterns, and tools have inspired aspects of our research. In particular, we drew conceptual inspiration from several public repositories and projects related to medical imaging, radiation therapy, and computer vision for healthcare applications.
We'd like to recognize the following projects which informed aspects of our approach and interface design.
Technology Stack
Body Models
Depth Estimation
Soft Tissue
We are grateful to all the researchers and developers who have made their work available to the community.