Case: Automating breast cancer treatment planning

When women are diagnosed with breast cancer and need to be treated with radiation, a dose treatment plan needs to be made to determine the right amount and incoming angles/location of dosage. This is necessary to obtain a conformal and homogeneous dose distribution, which provides enough dose in the tumor to destroy the malignant cells but also spares the nearby vital organs like lungs and heart. For her thesis Renée van Erp researched how the process of creating dose treatment plans for locally advanced left-sided breast cancer patients could be automated to be more efficient and less time consuming.

Background Pattern

The challenge

The process of creating dose treatment plans is a time consuming and iterative process. As seen in the treatment planning workflow a lot of steps, people and checks are involved before the ultimate dose treatment plan is ready. The dose treatments are also planner dependent due to differences in planning-experience, -skills or -training amongst treatment planners.

What she did

In her study Renée used a U-Net convolutional neural network, HD U-Net and Attention U-Net to create models that speed up the treatment planning process while maintaining quality consistency. Renée’s research explored the implementation of several data augmentation techniques and used a data-centric approach. Her work is the first in this field to inquire into the effect of data input channels, as it compares a 1-channel input versus a 5-channel data input. Her work used the structures of the planning target volume (PTV), optionally the organs at risk (OAR), and the clinically optimized dose distribution.

The result

The HD U-Net showed the best model performance on most evaluation metrics. Because of Renée’s research, the process of creating dose treatment plans is automated which can make it more efficient and less time consuming. The results of her research also show that the obtained dose treatment plans are clinically acceptable.

Interview with Master Thesis Award Winner Renée

Being a student in our Master’s Program Data Science in Business and Entrepreneurship gives you room to develop yourself to be an outstanding Data Scientist. The unique curriculum built with the support of Tilburg University and Eindhoven University of Technology gives plenty of room to students with innovative ambitions. But how do our students and Alumni experience the program? How do they shape it to fit their needs? Answers to these questions and more! We sat down with Renée van Erp, freshly graduated from the Master DSBE. She is also the winner of the Master Thesis Award 2023 of Tilburg University with her thesis entitled Breast cancer treatment planning: A U-net approach, she concluded her Master’s in Data Science and Entrepreneurship.

Click for the full article

Why partner up with JADS

JADS provides innovative data science programs at undergraduate, graduate and postgraduate level, conducts cutting edge data science research and offers great business opportunities with a continually growing ecosystem. From startups to SMEs and large corporates, our business partners gain access to talent, high level knowledge and business opportunities.

Partner up with JADS like 300+ other organizations and businesses and shape your data-driven future.

Learn more about business opportunities
Group 5
Group 6
Group 7