Postdoctoral Researcher

About the role

The University of Texas MD Anderson Cancer Center’s Department of Radiation Physics is seeking two postdoctoral fellows to develop BrachyGo, an AI-based treatment planning framework for cervical cancer brachytherapy. BrachyGo will be a major component of the broader Brachytherapy Planning Assistant (BPA), a web-based platform designed to streamline high-quality brachytherapy planning and expand accessibility in resource-limited settings.

The research focuses on patient-specific recurrence-risk prediction and reinforcement learning–based treatment planning.

Responsibilities

• Develop deep learning models that generate spatially resolved, three-dimensional recurrence-risk maps from multimodal clinical and imaging data.
• Use patient-specific risk maps to identify regions with elevated recurrence risk and inform spatial dose optimization.
• Develop reinforcement learning methods that learn and reproduce expert treatment-planning strategies.
• Develop and evaluate AI models for risk prediction and automated treatment planning.
• Compare AI-generated plans with expert clinical plans and investigate approaches that preserve physician interaction and individualized clinical decision-making.

Requirements

• Prior experience in machine learning, deep learning, reinforcement learning, or medical image analysis is strongly preferred.
• Experience with Python and common machine-learning frameworks such as PyTorch is desirable.

Key facts

• The role is to develop BrachyGo, an AI-based treatment planning framework for cervical cancer brachytherapy.
• The research focuses on patient-specific recurrence-risk prediction and reinforcement learning–based treatment planning.
• Experience in machine learning, deep learning, reinforcement learning, or medical image analysis is strongly preferred.
• Python and common machine-learning frameworks such as PyTorch are desirable.