Programme

Overview: The US government has launched the Cancer Moonshot program in 2016, aiming to “reduce the cancer death rate by half within 25 years and improve the lives of people with cancer and cancer survivors.” The program emphasis, that it “invests heavily in opportunities to speed delivery of cancer drugs and vaccines to prevent and treat cancer”. Radiation Oncology or Radiotherapy are not mentioned. In this debate, we will hear four proposals from four esteemed colleagues, describing a Cancer Moonshot program for Radiation Oncology, following ESTRO`s vision “Optimal Health for All”. The programs aim at transformative breakthroughs in Radiation Oncology, to substantially reduce cancer death rate and improve quality of life. The esteemed colleagues have a different professional background – one Radiation Oncologist, one medical Physicist, one Radiobiologist and one RTT. However, it is obvious that such an ambitious goal can only be achieved together, including our ESTRO society. All colleagues aim to convince the ESTRO conference participants to dedicate one Billion Euros over a period of 15 years to realize their vision of a Radiation Oncology Cancer Moonshot program.

Overview: Artificial intelligence is transforming radiation oncology, driving advances in fields such as medical imaging, tumor segmentation, and decision support. This session brings together experts at the forefront of AI applications, showcasing the latest innovations in these fields. Kareem Abdul Wahid will open the session with insights into how data science competitions accelerate AI advancements in radiation oncology, with a focus on the HNTS-MRG 2024 Challenge for MRI-guided tumor segmentation. He will discuss how benchmarking competitions accelerate the development of deep learning models, highlighting key takeaways on architecture choices, ensembling strategies, and training methodologies. Matteo Maspero will then explore the role of physics-informed neural networks (PINNs) in synthetic CT (sCT) generation from MRI. He will explain how incorporating physical priors enhances AI model performance, discuss applications in quantitative MRI and radiotherapy planning, and examine future potential for expanding physics-informed models in medical imaging. Expanding the discussion to broader applications, Ana Maria Barragan Montero will provide a comprehensive overview of foundation models in radiotherapy, discussing their advantages over traditional architectures like UNet. She will explore applications in medical imaging, multimodal models, and address critical challenges in model trustworthiness, reliability, and sustainability. Finally, Florian Putz will showcase the emerging role of large language models (LLMs) in radiation oncology, from decision support and automated documentation to multimodal AI-powered research tools. He will highlight LLM integration with oncology information systems, workflow automation, and privacy-preserving AI solutions for clinical applications. This symposium provides a cohesive journey from data-driven AI innovations to foundational AI principles and real-world clinical applications. Attendees will gain a comprehensive understanding of how the latest AI advancements are transforming radiation oncology and shaping its future.