Session

Sunday
November 29
08:45 - 10:00
Physics Stream 2
Application of machine learning to CTV definition
CTV definition is particularly complex as it accounts for microscopic tumor extensions beyond the GTV that are not visible with current imaging techniques. Machine learning is useful to automate current guidelines, increase consistency in delineation and ultimately improve the definition of CTV. This symposium will present three complementary views of data driven CTV definition. In a first part, Prof. Friedl will focus on the biological aspect of tumor progression based on the analysis of pathological tissue and outcome after intervention. A second part presented by Prof. Aristophanous will review CTV segmentation and auto-delineation techniques based radiological images with applications in head and neck cancer. In the third part, Prof. Unkelbach will discuss phenomenological models of tumor progression to support CTV definition for glioblastoma and head and neck cancer.
2130
Symposium
Physics
08:45 - 09:10
09:35 - 10:00