ESTRO 38 Abstract book

S361 ESTRO 38

improved for regions suffering from metal artefacts in conventional CT reconstruction, suggesting a direct benefit for delineation purposes. However, as these algorithms can also quantitatively alter the image in regions not influenced by metal artefacts, great care should be taken – especially in particle therapy planning. In summary, state-of-the-art CT imaging can provide additional value for radiation oncology purposes. This talk aims to increase awareness of this potential. By revisiting the institutional imaging protocol, one can potentially improve the image quality for delineation and/or safe dose to patients. SP-0699 Development of MR techniques focused on improved delineation M. Philippens 1 1 UMC Utrecht, Department of Radiation Oncology- Q 00.3.11, Utrecht, The Netherlands Abstract text Tumor delineation and delineation of organs at risk is an important but still difficult step in steering the dose in the treatment planning. Contouring of the tumor is mostly guided by changes in anatomy, asymmetry, and contrasts in the images. Also for the delineation of OAR contrast is needed to distinguish the organ from the background. As CT contrast of soft tissues is limited even after contrast administration, MRI is a logical choice to improve contrasts. The MR signal originates from the protons in water or fat. By manipulating the signal in a MR sequence the contrast can be adjusted. The delineation of gross tumor tissue (GTV) is in conventional radiotherapy mostly restricted to the pretreatment phase or when large changes have been recognized during therapy. However, with the introduction of MR linacs in the radiotherapy, MR delineation of both the GTV and OAR will become daily practice during treatment. Conventional MR contrasts are T1 and T2 weighted (T1w and T2w, respectively), which can be combined with fat suppression as fat has a high signal intensity on both T1 weighted as T2 weighted MRI. Generally, observers do agree better on delineation of OAR on MRI than on CT. Also for prostate and nasopharynx tumors, MRI has shown to decrease the target volume. For other tumor sites, increase of tumor volume is very common. This might be due to tumor induced changes in the vicinity of the tumor, which are included by the observers. This shows that although the soft contrast is much better on MRI, the interpretation might be more difficult. Therefore, clear guidelines are needed to improve tumor delineation or MR contrasts which are easier to interpret. Both dynamic contrast MRI and diffusion weighted MRI are widely used. These techniques, also referred to as functional imaging, reflect the status of the microvasculature and the microanatomy. The drawback of DCE-MRI is that rigorous post-processing is needed to effectively use it for contouring. Diffusion weighted MRI gives high contrast, but is frequently inadequate due to deformations which are inherent to the acquisition method used, echo planar imaging (EPI). Recently, several advances have been made to mitigate these distortions. One approach is to decrease the distortions in the EPI using acquisition acceleration methods. Another approach is to replace the EPI acquisition with a fast spin echo or steady state free precession method, which are both used for standard clinical MR acquisition. The disadvantage is that the acquisition is slower and motion artefacts have to be tackled. The role of DCE-MRI in delineating the gross tumor volume (GTV) has been studied in prostate, head and neck, cervical cancer and brain. Diffusion weighted MRI has been more frequently studied as a tool to improve the delineation. The high contrast between tumor and background might even allow automatic delineation. Also diffusion tensor imaging (DTI)

1 Mirada Medical Limited, Science and Medical Technology, Oxford, United Kingdom

Abstract text “You don't learn to walk by following rules. You learn by doing, and by falling over” – Richard Branson. To learn from our mistakes, we need to understand that we have made them first, and then find ways to address them. In this talk, we will look at the ways through which we can assess contouring accuracy, including quantitative, qualitative and dosimetric approaches, giving examples from published literature. However, such assessment in often confounded by clinical inter-observer variation, so we will also consider how we can address this is our assessments and consider its potential clinical implications. We will summarise by discussing what approaches can be taken to help us improve contouring quality. SP-0698 CT-based delineation: What can we gain from state-of-the-art CT image acquisition and reconstruction techniques C. Richter 1,2,3,4 , F. Negwer 1 , E.C. Troost 1,2,3,4,5 , P. Wohlfahrt 1,2 1 Oncoray - National Center For Radiation Research in Oncology, Faculty of Medicine and University Hospital Carl Gustav Carus- Technische Universität Dresden- Helmholtz-Zentrum Dresden - Rossendorf- Dresden- Germany, Dresden, Germany; 2 Helmholtz-Zentrum Dresden-Rossendorf, Institute Of Radiooncology, Dresden, Germany ; 3 Department of Radiotherapy and Radiation Oncology, Faculty Of Medicine and University Hospital Carl Gustav Carus- Technische Universität Dresden, Dresden, Germany ; 4 german Cancer Consortium Dktk, Partner Site Dresden, Dresden, Germany ; 5 national Center for Tumor Diseases NCT, Partner Site Nresden, Dresden, Germany Abstract text X-ray computed tomography (CT) has been the standard imaging modality in radiation oncology for both, treatment planning and delineation of targets and organs at risk for decades. For further improvement, especially for delineation, magnetic resonance imaging (MRI) and positron emission tomography (PET) are being extensively investigated and more often included into clinical routines. They can provide better soft tissue contrast and functional information. Still, also in the field of CT imaging relevant improvements have been made, that are not so much in the spotlight. Hence, this talk will focus on novel CT image acquisition and reconstruction techniques and their potential benefit for radiation oncology applications. First, the potential value of dual-energy CT (DECT) for delineation will be discussed. DECT has already been proven to allow for a more accurate treatment planning, especially in particle therapy. It provides additional tissue information compared to conventional CT imaging. Furthermore, DECT enables the reconstruction of different CT datasets with varying image contrasts. Currently, it is unproven whether this additional information translates into improvement of the segmentation and delineation quality. The exploration of this benefit in combination with machine learning approaches is envisioned. First studies will be presented. Second, the potential of iterative CT reconstruction methods will be highlighted. They allow for a substantial reduction of imaging dose to reach a similar noise level as conventional filtered back projection. Hence, iterative reconstruction is of high relevance for adaptive protocols as it reduces the dose burden from more frequent CT imaging during treatment. Third, the value and challenges of metal artefact reduction algorithms will be covered. It has been shown that the visual image impression can be substantially