S2981
Invited Speaker
ESTRO 2026
imaging information into clinically meaningful parameters.
and therapeutic index. Imaging is nowadays ubiquitous across the treatment pathway from disease diagnosis and treatment selection to planning, delivery, and follow-up. In this talk we will explore how novel developments in multimodal imaging and image processing can address current shortfalls in radiation oncology workflows. Medical imaging technology development should aim to capture, at the right time, the relevant information required for personalised clinical decisions. This information is diverse – i.e. morphological, physical, functional and biological characteristics of the tumour and normal tissues – highlighting the importance of multimodal approaches to monitor these complex processes. There is a need to refine established modalities – such as CT, MRI, and PET – towards higher spatial and temporal resolution, novel acquisitions, increased quantitative capability, reduced imaging dose, acquisition standardisation, shorter acquisition times and integration into delivery systems, as well as translating emerging modalities like particle imaging into clinical routine. Routinely used modalities also contain a wealth of information that is currently underutilised. For example, dual-energy CT offers rich morphological and physical information but, despite increasing availability, it remains poorly integrated. Novel biomarkers from CT, MRI and/or PET linked to histopathological and molecular data, may hold prognostic value. Widely available techniques such as radiographs and surface imaging may also yield richer information when combined with advanced analysis methods. Finally, currently clinical decisions remain largely based on “snapshot” imaging information acquired at specific periods in time. In contrast, both the tumour and the healthy tissues are inherently dynamic, evolving due to disease progression and in response to treatment. Looking ahead, there is a need for a paradigm shift toward patient-specific digital twins – i.e., longitudinal models in which the morphological, physical, functional and biological patterns are continuously updated through the integration of discrete imaging information and complementary data sources to infer prognostic information on disease progression, treatment response and toxicity. Such an approach would enable a transition toward predictive radiotherapy workflows, supporting fully adaptive and personalised treatment strategies. The future of imaging in radiation oncology therefore lies on the development of novel technologies and their effective integration and exploitation. Leveraging imaging information as a key input into longitudinal, data-driven frameworks will be key to maximising the targeting capabilities and therapeutic index of radiation treatments. Medical physicists will continue to have a key role in measuring and converting
5256 Particle: Robust treatment planning Edmond Sterpin Oncology, KU Leuven, Leuven, Belgium. MIRO lab, UCLouvain, Brussels, Belgium In photon radiotherapy, the limitations of the static dose cloud approximation have been recognized for many years, particularly in well-documented clinical situations such as breast irradiation and the treatment of small lung tumors. In these contexts, anatomical changes, motion, and setup uncertainties can significantly compromise the validity of conventional planning assumptions. To mitigate these effects, several intermediate strategies have been explored within the photon community, including approaches such as flash irradiation for skin dose adjustments or gross tumor volume–based dose prescription. While these methods have provided partial improvements, they have not delivered a truly comprehensive or satisfactory solution. In particular, they fail to fully address the fundamental issue of the interplay between dose distributions and uncertainties. In proton therapy, by contrast, the failure of the static dose cloud approximation is not an exception but rather the rule, due to the strong sensitivity of proton dose distributions to range uncertainties and anatomical variations. As a result, robust optimization rapidly emerged as a clinical necessity and is now routinely implemented in proton treatment planning. In this presentation, I will illustrate how robust optimization has reshaped clinical proton therapy practice. More recently, the photon therapy community has begun to translate robust optimization concepts into photon treatment planning, acknowledging that similar uncertainty-related challenges persist, albeit through different physical mechanisms. Emerging clinical and research examples will be discussed to highlight this ongoing transition and its potential implications for photon radiotherapy workflows. This talk aims to foster cross-fertilization between proton and photon communities around robustness concepts. In this spirit, participants with practical experience in robust optimization applied to photon therapy are warmly invited to share their insights and perspectives.
Made with FlippingBook - Share PDF online