ESTRO 2025 - Abstract Book

S718

Clinical - CNS

ESTRO 2025

Conclusion: This study demonstrates that black blood MRI (MSDE) provides superior interobserver agreement compared to standard 1C imaging for brain metastasis contouring, as shown by improved DSC values. By reducing interobserver variability, MSDE enhances confidence in volume delineation and supports consistent treatment planning. The findings suggest integrating black blood MRI into routine workflows to optimise contouring precision and improve the quality of radiotherapy outcomes.

Keywords: contouring, MRI-guided, interobserver-agreement

3146

Poster Discussion 7-Tesla vs. 3-Tesla MRI in low grade glioma radiotherapy planning: Can we improve delineation consistency? Reinhardt Krcek 1,2 , Robert Poel 1,3 , Nicolas Bachmann 1 , Hossein Hemmatazad 1 , Philipp Reinhardt 1 , Piotr Radojewski 4,5 , Johannes Slotboom 4,5 , Mauricio Reyes 3 , Ekin Ermiş 1 1 Department of Radiation Oncology, Inselspital, Bern University Hospital and University of Bern, Bern, Switzerland. 2 Radiation Medicine Program, Princess Margaret Cancer Centre, University Health Network, Toronto, Canada. 3 ARTORG Center for Biomedical Research, University of Bern, Bern, Switzerland. 4 Institute of Diagnostic and Interventional Neuroradiology, Bern University Hospital and University of Bern, Bern, Switzerland. 5 Translational Imaging Center (TIC), Swiss Institute for Translational and Entrepreneurial Medicine, Bern, Switzerland Purpose/Objective: Low-grade gliomas represent approximately 2/3 of gliomas in young adults. Surgery and radiotherapy (RT) are the primary local treatment modalities. Given the relatively young age of these patients, optimal imaging for RT planning is desirable. Currently, the standard imaging for low-grade gliomas is based on 1.5 Tesla or 3 Tesla MRI (3T MR). The availability of ultra-high-field 7 Tesla MRI (7T MR) offers enhanced resolution and an improved signal-to-noise ratio (1). To evaluate the potential advantages of 7T MR for RT planning in low-grade gliomas, we assessed inter- and intra-rater variability in the contouring of target volumes and OARs using 7T MR compared to 3T MR. Material/Methods: Thirty patients with suspected low-grade gliomas underwent non-contrast 3T MR (Siemens Magnetom Prisma, Magnetom Vida, Skyra Fit) and 7T MR (Siemens Magnetom Terra) before surgery. The 3T MR scans included T1, T2, and FLAIR sequences, while the 7T MR scans included T1 and T2 for all, and FLAIR sequences for most patients. Three experienced radiation oncologists delineated OARs and target volumes according to international guidelines. Volume information was collected and the Dice coefficient (DC), a measure of volumetric overlap and error, was used to assess inter-rater variability. Intra-rater variability was analyzed by comparing the DC of each 3T/7T MR pair per rater. Results: A total of 3,575 structures were analyzed. For the GTV, the mean DC between raters was 0.84 (± 0.13) for 3T MR and 0.81 (± 0.16) for 7T MR. The mean DC for CTV was 0.90 (± 0.05) on 3T MR and 0.89 (± 0.07) on 7T MR. OARs showed a mean DC of 0.74 (± 0.15) for 3T and 0.72 (± 0.17) for 7T. For intra-rater variability between 3T and 7T MR, the mean DC for target volumes was 0.88 (± 0.09), while for OARs it was 0.66 (± 0.18). The mean GTV volume was 38.9 cc (± 37.9) for 3T and 36.8 cc (± 36.1) for 7T. Corresponding PTV volumes were 222.4 cc (± 133.5) for 3T and 207.0 cc (± 126.5) for 7T.