Applications
Clinical settings that motivate the research.
- Cancer Early Screening
- AI-driven Treatment Planning Systems
- Image-Guided Radiotherapy
- Upright Proton Therapy
Research Introduction
MedAPT (Medical AI for Precision Therapy) develops medical AI methods that connect data-driven generative models with physical principles to solve difficult medical inverse problems. Our goal is to make precision cancer care—especially proton therapy—more accurate, efficient, and accessible.
Research Framework
Our work follows a connected path from clinical applications, through core imaging tasks, to algorithms that make the pathway possible.
Tasks
Medical imaging capabilities that support precision therapy.
- Image Reconstruction
- Image Generation
- Image Registration
- Image Segmentation
- 4D Motion Analysis
Algorithms
Learning and physical modeling tools that connect data to care.
- Physics-Informed Neural Networks
- 3D Gaussian Splatting
- Flow-Matching Generative Models
- Foundational Medical Models
Explore by Topic
Medical Imaging
Precision Radiotherapy
Proton Therapy
Deformable Image Registration
MR-to-CT Synthesis
Dose Calculation
Treatment Planning
Motion Modeling
4D Imaging
Image Reconstruction
Uncertainty Estimation
Synthetic CT
Brain Tumor Imaging
Head-and-Neck Cancer
Diffusion Models
Generative Models
Gaussian Representation
Semantic Segmentation
Computer Vision
Machine Learning
Drag · click a topic