ESTRO 2022 - Abstract Book

S1434

Abstract book

ESTRO 2022

Conclusion As a first step towards the development of an optimization tool for MR sequences, the DL-based model maintained a higher prediction accuracy for both SNR and CNR in the validation data set. The next step is to develop and implement multi- objective optimization methods with regards to SNR and CNR based regression.

PO-1639 A pre-treatment patient specific QA alternative approach for HyperArc treatments.

N. Cavalli 1 , E. Bonanno 2 , G. Stella 3 , G. Borzì 1 , L. Zirone 1 , M. Pace 1 , C. Marino 1

1 Humanitas-Istituto Clinico Catanese, Medical Physics Department, Misterbianco, Italy; 2 Humanitas-Istituto Clinico Catanese, Medical Physics Department, Misterbianco , Italy; 3 University of Catania, Physics and Astronomy "E. Majorana" Department, Catania, Italy Purpose or Objective Mobius3D system was validated with the purpose to use it as a modern second – check dosimetry system. Our objective has been to determine the appropriateness of M3D as pre-treatment patient specific QA for HyperArc (HA) treatments, establishing if Mobius3D can safely replace the measurements based patient specific QA for this kind of treatments. Materials and Methods 30 Stereotactic Radiosurgey (SRS) HA plans for brain were delivered across the ionization chamber located inside homogeneous Mobius Verification Phantom (MVP). The measured dose inside the MVP was compared with M3D calculated dose and then with TPS calculated one. In a second step for all 30 plans, dose distributions calculated on the patient’s CT- dataset by Mobius and our algorithm ( AcurosXB v.15.6 ) were compared using the following metrics: target D mean percent difference, target D 90% percent difference and global 3D gamma passing rate over the entire dataset. The used gamma criteria were 3%-2mm. All 30 plans were then delivered across radiochromic film (EBT3) and electronic portal imaging device (EPID) and analysed with Portal Dosimetry (PD), in terms of gamma analysis (3%-2mm). Results were evaluated considering a gamma passing rate >90 %. The appropriateness of M3D as an alternative to measurement-based patient specific QA, was evaluated as the percentage of verified plans that passing 3D global gamma analysis with Mobius3D and the conventional methods at the indicated level. Results All 30 analyzed plans showed good agreement between MVP measured point dose and dose calculated by M3D and Eclipse TPS. In fact, the average point dose difference on MVP between our algorithms and measurements with ionization chamber was 1.85 % [min 1.16%-max 2.55%], while concerning M3D, the average point dose difference was 2.32 % [min 1.05%-max 3.44%]. Concerning comparison on the CT-dataset between TPS and Mobius3D we found an average target D mean percent difference of 1.83% [min 1.16%-max 2.55%]. The average D 90% difference was 2.83% [min 2.6%-max 3.1%]; 3D global gamma passing rate ranged from 89.2% to 99.6 % at the specified gamma criteria.