ESTRO 2024 - Abstract Book

S130

Invited Speaker

ESTRO 2024

3548

Virtual patients for clinical trials

Ehsan Abadi

Duke University, Radiology, Durham, USA

Abstract:

A successful virtual clinical trial hinges on use of virtual patients that accurately represent the attributes needed to address the question of the trial. In this presentation, we discuss the evolution and significance of virtual patients over the years. We cover various approaches on modeling anatomy, diseases, and physiology. We discuss diversity and variability that is needed to model a targeted population. The talk includes some examples on the clinical utility of patient models in virtual trials. Additionally, we touch on essential requirements for widespread use of virtual patients including considerations of credibility, reproducibility, informatics, and accessibility. The presentation highlights some prospects of virtual patients and their crucial role in precision medicine.

3549

What are the major bottlenecks for getting it into clinics?

Edmond Sterpin

KU Leuven, Oncology, Leuven, Belgium. UCLouvain, MIRO, Brussels, Belgium. PARTICLE, PARTICLE, Leuven, Belgium

Abstract:

Online adaptive particle therapy requires for its effective clinical implementation the rapid execution of acquisition and processing of complex data, without compromise with regard to the quality of the treatment plan and patient safety. Very quickly, the clinical workflow can be summarized as follows: 1) acquisition of a volumetric image; 2) patient positioning; 3) image segmentation using (or not) the initial planning data; 4) conversion of the image into physical data for dose calculation; 5) optimization (preferably robust); 6) quality assurance; 7) delivery of treatment The second point generally does not pose a specific problem. The third point is crucial but the problem is shared with X-ray radiotherapy. It must be remembered that in existing commercial X-ray solutions based on CBCT imaging, the segmentation of the image into contours is the point that takes the most time in the workflow. All the other points, however, have aspects specific to particle therapy, and these points are often major. For example, the quality of onboard imaging must follow stricter specifications in particle therapy because dose calculation requires precise physical data that current CBCTs are not able to provide natively (synthetic CT generation strategies must be implemented, with their limitations). We can temporarily fall back on on-rails CT systems, but the latter are less available and also have their limits in terms of integration and execution speed. Treatment optimization reveals an additional difficulty in particle therapy in case of robust optimization, which takes considerably more time than traditional optimization with a PTV. Finally, quality control must take into account the specificities of particle therapy