ESTRO 2021 Abstract Book

S1664

ESTRO 2021

supplied by the vendor. The third sequence was optimised using an anatomical phantom with surrogate titanium implants, a higher receive bandwidth and adjusted echo times and spacing to mimic a metal artefact reduction sequence. Prior to evaluation in patients, images of a large field image distortion phantom (CIRS) were acquired to ascertain true geometric distortion in sequences described above. MR images were rigidly registered to the pCT by a single observer using bony anatomy. The pCT was used as the reference “non-distorted” anatomical standard. Coronal and axial images bisecting the treatment isocentre were selected and bony landmarks identified. 4 of 6 landmarks on axial imaging had corresponding landmarks in the coronal plane (Figure 1). Using the fixed image registrations, 5 observers placed points of interest on the pCT then the MRI. The difference between landmark coordinates on the pCT and MRI was taken to be the measure of perceived distortion within the MR image. To avoid bias from the varying appearance of cortical bone between the two modalities, training was given to all observers. Results Results of CIRS phantom demonstrated geometric distortion was <1.0mm for all three sequences at distances <15cm from the imaging isocentre. Of the 4 patients included, 3 had prosthesis and 1 (the control) did not. Table 1 illustrates the mean difference in position between the bony landmark coordinates on pCT and MRI across 5 observers. Overall, mean difference was <5 mm, and ≤ 3mm in 88% of measured distances. One-way analysis of variance between observer means showed no statistically significant difference in the distortion between sequences and intra-class correlation between observers showed moderate reliability on this very small

cohort. Conclusion

This in-vivo feasibility study of the impact of metallic prostheses on MR images acquired on an MR Linac in a small cohort suggests that distortion due to metallic hip prosthesis may not be an obstacle for prostate MRgRT. Further investigation of the impact on dose distribution on a wider cohort of both patients and range of hip prosthesis materials is warranted.

Digital Poster: Treatment planning and dose calculation/QC and QA

PO-1953 Investigation of the effect of minimized tumor volumes on critical organ doses in GBM radiotherapy M.C. Şahin 1 , İ. Kaya 2 , İ.D. Cingöz 2 , E. Bozoğlan 1 , K. Manisa 3 , Z.A. Palabıyık 4 , N. Yüceer 5 1 Kutahya Health Sciences University, Training and Research Center, Kütahya, Turkey; 2 Usak University, Neurosurgery, Uşak, Turkey; 3 Kutahya Dumlupınar University, Nuclear Physics, Kütahya, Turkey; 4 Sakarya Training And Research Hospital, Radiation Oncology, Sakarya, Turkey; 5 Katip Çelebi University , Neurosurgery, Izmir, Turkey Purpose or Objective Glioblastoma multiforme (GBM) is the most common malignant brain tumor in adults. Its standard treatment is regional radiotherapy with simultaneous temozolomide chemotherapy after safe maximum surgical resection. The margins and high doses used in radiotherapy planning for GBM patients increase the potential for acute and late neurotoxicity. In this study, we aimed to compare the treatment plans made using cranial and tumor models specific to the patient’s anatomy produced using minimized tumor volumes (PTV min ) and a three-dimensional printer, and the treatment plans of patients, and to compare the tumor and critical organ doses and treatment times. Materials and Methods Ten patients who were diagnosed with GBM between June 2018 and March 2019 were included in our study. Surgical treatment included maximum safe resection of the tumor. External treatment plans of five-field sliding window intensity- modulated radiotherapy (IMRT) and two-half-arc volumetric modulated arc therapy (VMAT) treatment plans were prepared using 3D models created from PTV min volumes from 10 patients (Group 1). In the control group (Group 2), five-field sliding window IMRT and two-half-arc VMAT external treatment plans were prepared using routine computed tomography images from the same 10 patients for radiotherapy. Based on these treatment plans, tumor volumes and doses received by critical organs were compared. Results The dose values obtained from the contoured PTV original and minimized volumes for radiotherapy treatment were compared with each other; the dose values obtained from the treatment plans prepared for the contoured PTV min and minimized volumes were found to be statistically significantly different (p<0.05). Despite the differences in point dose data obtained from the treatment planning system, the quality control irradiations made using the Electronic Portal Imaging Device for all