Machine Learning

Our machine learning research team is developing intelligent tools to enable the next generation of highly personalized and effective radiation therapy. Our work focuses on automating and improving critical steps in the adaptive radiotherapy workflow.

Our Core Research Areas

1. AI-Powered Automatic Segmentation

Precision in radiotherapy begins with accurately identifying the tumor and nearby healthy organs (organs at risk). We are creating deep learning models that can automatically and rapidly delineate these structures from CT and MRI scans. This automation is crucial for adaptive radiotherapy, as it allows for quick plan adjustments in response to anatomical changes, ensuring treatment remains targeted and safe throughout the entire course of therapy.

Example of automatic segmentation of organs at risk on a CT scan..

2. Deformable Image Registration

A patient's anatomy can change significantly from one treatment session to the next. Our research in deformable image registration uses advanced AI algorithms to precisely align and map these changes over time. By tracking how a tumor shrinks or how internal organs move, our technology provides the essential data needed to adapt the treatment plan, ensuring the radiation dose is delivered to the correct location, every single time.

Visualization of a vector field representing anatomical changes

Our Vision

Our goal is to harness the power of AI to provide clinicians with the tools for faster and more accurate decision-making. By automating these complex processes, we aim to make truly adaptive, personalized radiotherapy a clinical reality, improving outcomes for patients.