MSc Thesis: Evaluating LLM-Based Agentic Systems in Radiotherapy Workflows

How well can AI agents perform tasks in radiotherapy treatment planning? In this master’s thesis, you will develop and evaluate agentic systems using different LLMs and radiotherapy software.

Project Description 

The thesis will investigate how LLM-based agentic systems can be used for tasks in the radiotherapy workflow. Radiotherapy treatment planning is a complex process with several steps, from patient intake and prescription to patient modelling, plan creation and plan review.

The work will explore how different LLMs, agent harnesses and tools affect the ability to complete these tasks. A small benchmark will be developed to evaluate and compare the approaches. Data, test cases and the radiotherapy software needed for the project will be provided.

Your main tasks

• Review research on medical agent benchmarks and LLM-based agentic treatment planning.
• Define research questions and evaluation criteria.
• Create a benchmark of treatment planning tasks.
• Develop agent harnesses and compare them with existing open-source and proprietary harnesses.
• Evaluate different LLMs, harnesses and tool setups using the benchmark.
• Write and present the thesis.

Your profile

We are looking for a curious master's student who wants to work where AI meets cancer care. You enjoy combining programming with careful experiments, and you look critically at results instead of taking them at face value. You are independent and well organized, and you can explain your findings clearly to both software engineers and medical physicists.

• Ongoing master's studies in engineering physics, medical physics, computer science, applied mathematics or a related field
• Coursework or project experience in optimization and/or machine learning
• Great programming skills in Python
• Hands-on experience with LLMs
• Fluent written and spoken English

Application

Please apply to the position through the link below. Selection and interviews will be ongoing. We do not accept applications by e-mail. 

Application Link.