ESTRO 2020 Abstract Book

S798 ESTRO 2020

Results

best AI-driven IMRT plans and the reference plans from the standard TPS (Table 1). The CI was similar between the two TPS for IMRT plans and better for VMAT reference plan compared to the AI driven plan. The HI was closer to 0 for the reference plan for both IMRT and VMAT plan. The modulation complexity was higher for the AI-driven VMAT plans compared to the reference plan and comparable for the IMRT treatment plans generated by the two TPS. The dose to the normal tissue was equivalent for all techniques investigated (Table 2). Conclusion At treatment planning level, the study demonstrated that the use of an AI-driven TPS are competitive with the current IMRT/VMAT plans created with a standard TPS and can improve the sparing of healthy tissues while maintaining a full coverage of the PTV. Further studies are required to confirm these early results. PO-1484 MRI only based planning to decrease toxicity in prostate cancer IMRT/IGRT: a dosimetric study K. Stephane 1 , A. Houssayni 2 , U. Cassard 2 , A. Barateau 2 , L. Duvergé 1 , A. Largent 2 , O. Acosta 2 , J. Nunes 2 , T. Messai 1 , A. Simon 2 , J. Dowling 3 , N. Perichon 1 , R. De Crevoisier 1 1 Centre Eugene Marquis, Radiotherapy, Rennes, France ; 2 University Rennes, Inserm LTSI-UMR 1099, Rennes, France ; 3 CSIRO, e-Health Research Centre, Herston, Australia Purpose or Objective Unlike CT-scans, MRI allows improved visualization of the prostate and the anatomical structures involved in erectile function. The aim of this study was to evaluate the dosimetric benefit of MRI-only based radiotherapy planning to spare the organs-at-risk (OARs) in IMRT/IGRT for prostate cancer, compared to a standard radiotherapy planning from CT. Material and Methods A total of 15 patients were considered in this study. Each patient had a CT and an MRI scan. Pseudo-CTs (pCT) were generated from MRI scans using a generative adversarial network deep learning method. “Standard” volumes of interest such as prostate, seminal vesicles (SV), rectum, bladder and femoral heads were manually delineated on pCT (to mimic contours on CT-scan) and MR images. However, penile bulb (PB), internal pudendal arteries (IPA) and corpus cavernosum (CC) were only delineated on MRI. Then, two planning strategies were computed on the pCT for each patient, with the Pinnacle (Philips) treatment planning system using a VMAT technique (6 MV, 1 arc per phase). One strategy used the contours from the pCT and the other one the contours from the MRI, both delivering 78 Gy in the prostate and 50 Gy in the SV. The CT-based planning considered dose constraints for the “standard” OARs, respecting the GETUG/RTOG recommendations. The MRI-based planning considered dose constraints in the “standard” OARs and additional constraints to spare the PB, the IPA and the CC. The prostate volumes, DVHs, monitor units (MU), conformity and homogeneity indexes were compared between the two planning strategies using the Wilcoxon test. Results The prostate volume was significantly smaller using the contours from MRI (median value: 49.9 cc (range, 16.1- 121.6) than on CT (64.9 cc (32.4-124.2)) (p<0.01). The prostate coverage (V95%) was not significantly different between the two plans (98.6% versus 98.4%; p=0.10). Table 1 displays the dose in the OARs, by planning strategy. The doses were significantly reduced in all the OARs (except for the femoral heads) using the MRI-based planning. The number of MUs, conformity and homogeneity indexes were not significantly different between the two planning strategies

The IMRT plans generated by the AI-driven TPS were better at fulfilling the clinical goals than the generated VMAT plans for all the treatment sites. Both, the coverage of the PTV and the sparing of OAR were very similar between the