ESTRO 2021 Abstract Book

S418

ESTRO 2021

Results The mean cell density was highest in the posterior prostate region, within the PZ, with values reaching approximately 35000 cells/mm³, and lowest in the region of the urethra. Tumour lesion probability was also highest within the PZ, reaching approximately 40%, and displayed left/right symmetry (Figure 2). The biological atlas can be further stratified by tumour grade and patient-specific measurements such as age, prostate volume, and PSA level.

Conclusion The population-based biological atlas provides a quantitative, 3D model of the location and characteristics of prostate cancer in the study cohort. The novel use of intermediate ex vivo MRI reduces registration uncertainty caused by tissue deformation after prostate resection. There is an increasing role for mpMRI in the diagnosis and treatment of prostate cancer. Machine learning models have been developed to predict the location of prostate cancer lesions, and previous studies have demonstrated the potential for population-based tumour prevalence maps to substantially enhance model performance. We extend this idea by incorporating additional biological characteristics relevant for optimising quantitative biologically-guided radiation therapy, such as tumour grade and tumour cell density, generated by combining total cell density and tumour probability maps. A registration framework has been designed to map the atlas to in vivo MRI, to apply in machine learning models and to facilitate biologically-optimised treatment planning. OC-0532 IV contrast-enhanced CBCT position verification for SBRT of ventricular tachycardia. C. van Gijlswijk 1 , E. Astreinidou 1 , M. de Ridder 1 , K. Zeppenfeld 2 , M. Hol 1 , C. Rasch 1 1 Leiden University Medical Centre, LUMC, Radiotherapy, Leiden, The Netherlands; 2 Leiden University Medical Centre, LUMC, Cardiology, Leiden, The Netherlands Purpose or Objective To investigate the feasibility and the potential improvement of CBCT based position verification by using intravenous (IV) contrast, in patients treated with the novel technique of stereotactic radioablation for ventricular tachycardia (VT). Materials and Methods Three patients with VT, for which catheter ablation was not effective due to a deep anterior-basal VT substrate underwent treatment with a single stereotactic radiotherapy fraction of 25Gy. The PTV delineation, including left ventricle (LV) summit region, was based on converting electrophysiological maps of the heart to the treatment planning respiratory motion 4DCT. Two full coplanar VMAT arcs were used. Total treatment time was 45 min. Position verification was performed with 4DCBCT. Four CBCTs were performed: Set-up (A) correction, set-up verification (B), in-between the two arcs correction (C) and post treatment (D). Dual match method was used. At first bone match and then grey value match of a mask around the PTV +2 cm. In two Proffered papers: Proffered papers 34: Image-guided radiotherapy and treatment verification