ESTRO 2022 - Abstract Book

S407

Abstract book

ESTRO 2022

• Match algorithms; Bone/Grey value/Seeds/Mask/Manual. • Determining match structures regarding target area. • Protocol for anatomical/pathological changes.

Departments were asked about future developments regarding new techniques and specifically if they intended to use Varian Ethos or Magnetic Resonance Image guided Radiation Therapy (MRIgRT). Target areas (in the progress of) using these techniques were registered. All survey results were anonymized and compared with the current curriculum. Based on this, areas in need of improvement were identified. Results A total of 18 (86%) respondents completed the survey (one survey was incomplete). The most commonly used imaging technique was 3D CBCT, while 2D MV (Portal) Imaging was barely used (Figure 1). All respondents used online setup protocols, none used a fully offline setup protocol, and 28% (n=5) used a combined online/offline protocol. Regarding adaptive strategies, LOP was used or intended to be used for the bladder (55%, n=10), cervix (33%, n=6) and rectum (11%, n=2). About 39% (n=7) of departments intend to use MRIgRT and 11% (n=2) the Varian Ethos in the future. Every department had a protocol in place for observed anatomical changes, of which 83% (n=15) used the traffic light protocol. Except for MRIgRT, all techniques clinically being used were present in the curriculum. However, the traffic light protocol and LOP were only present to a limited extent.

Conclusion The IGRT curriculum is broadly in line with the current professional practice in the Netherlands. However, some adjustments are advisable. 2D MV (Portal) Imaging and offline setup protocols can become a smaller part of the IGRT curriculum as these techniques are no longer frequently used. The use of protocols dealing with anatomical changes, as well as the use of online adaptive strategies (LOP/MRIgRT), should be included and more prominent in a future curriculum.

OC-0462 An artificial intelligence approach to tumor volume delineation in glioblastoma

M. Hannisdal 1

1 Haukeland University Hospital, Department of Oncology and Medical Physics, Bergen, Norway

Purpose or Objective For glioblastoma (GBM), an isotropic margin of 20 mm from the contrast enhancing tumor is used because (i) conventional MRI and interpretation methods have limitations as to specify the location of non-enhancing infiltrative GBM, and (ii) recurrence commonly occur within 20 mm from the primary tumor. However, inter- and intra-observer variability and the diffuse radiological signature of GBM adds to the challenge of optimal visual interpretation and accurate delineation of GBM. A precise method is needed to specify the radiotherapy target and avoid unnecessary patient neurocognitive side effects. The objective of this study was to investigate if a deep learning model for segmenting GBM on multispectral MRI correlated to manual delineations, and thereby potentially be feasible as an oncologist support tool. Secondly, we aim to investigate if the 20mm margin is justifiable in respect of geometrical appearance of recurrent GBM harboring unmethylated O 6 - methylguanine-DNA methyltransferase (MGMT) promotor. Materials and Methods Longitudinal image sets of multispectral post-operative MRI from six patients with (i) primary- and (ii) recurrent GBM were analyzed. All patients harbored unmethylated MGMT promoter, indicating high radioresistance and poor prognosis. We used a deep learning pre-trained U-Net based algorithm called HD-GLIO, trained on 3220 GBM image sets labeled by neuro- radiologists. The enhancing core (automatic GTV) and non-enhancing GBM compartments (automatic CTV) were derived and compared to the clinical GTVs and CTVs, respectively. The image series of the primary tumor was rigidly co-registered with the image series of the recurrent GBM for volumetric comparison between the clinical CTV and the recurrent tumor.