Programme

Overview: In this session we welcome our two multidisciplinary speakers from The Catharina Cancer Institute in Eindhoven, The Netherlands. During this session we will have the opportunity to learn first-hand how deep learning segmentation (DLS) has been implemented in their clinic, and how this process has changed their practice. Firstly, Clinical Physicist, Coen Hurkmans, will consider the motivation to implement DLS practices in a clinical environment. Drawing on evidence from the literature, he will highlight achievable results, common pitfalls, and strategies to mitigate errors in segmentation. He will also provide us with a comprehensive guide to evaluating the quality of segmentations within clinical workflows. Coen will also address the how to establish a multidisciplinary team to support the implementation phase of DLS. His talk will then be followed up by RTT colleague, Melissa Verdonk-van den Heuvel. Melissa will talk us through the practicalities of introducing DLS, using examples from various imaging technologies, including CBCT and MRI. She will also highlight the outcomes of testing protocols, and the ongoing quality assurance measures in place critical to the success of DLS. Finally, Melissa will discuss the evolving roles of radiation therapists (RTTs) in this changing landscape and explore potential future opportunities for advanced practice.

Overview: This session will provide an overview of new applications and explore the future of AI in clinical practice from the perspective of both radiation oncologists and radiologists. The introduction of AI is revolutionising radiotherapy workflows by optimising tumour and organ segmentation, saving valuable time for radiation oncologists. In addition, AI has great potential in areas such as quality control, treatment plan verification, optimisation of image-guided radiotherapy and real-time monitoring of moving tumours, such as lung and liver tumours, during treatment. In diagnostic imaging, AI applications are emerging in screening programmes, prognostic assessment and predictive modelling. These areas are becoming more prominent in research, with early impacts already visible in clinical practice. The integration of AI-based models will be a key challenge in the coming years. The presentations will focus on the feasibility, testing and planning of innovative AI solutions in both radiotherapy and diagnostic imaging, emphasising the growing influence of AI on the workflows of oncologists and radiologists.