ESTRO 2024 - Abstract Book

S3733

Physics - Dose prediction, optimisation and applications of photon and electron planning

ESTRO 2024

In both hospitals participating in this study, significant resources are dedicated to monitor and reevaluate treatment plans in H&N cancer patients. Despite these efforts, dosimetric deviations higher than desired have been found in some cases. This justifies the need of a more automated tool to evaluate daily delivered doses for an evidence-based decision-making. A work in progress based on the large dataset described in this study (which for the portal images go beyond the gamma passing rates) suggests that dose predictive models based on artificial intelligence can be developed, which would reduce the workload and allow a more optimal identification of the treatments to be replanned.

Keywords: dose changes, head and neck, correlation

3131

Digital Poster

Development of an automated VMAT-technique for left locoregional breast radiotherapy including IMN

Jomar Frengen 1 , Marit Funderud 1 , Johan Vikstrøm 2 , Ingvil Mjaaland 2 , Sigrun Saur Almberg 1 , Kjell Ivar Dybvik 2 , Mari Helene Blihovde Hjelstuen 2 1 St. Olavs Hospital, Department of Radiotherapy, Trondheim, Norway. 2 Stavanger University Hospital, Department of Radiotherapy, Stavanger, Norway

Purpose/Objective:

Conventional three-dimensional conformal radiotherapy (3D-CRT) has been the preferred technique for breast radiotherapy for many years. For complex cases, the use of Volumetric Modulated Arc Therapy (VMAT) is increasing. However, the optimization process is usually not consistent and standardized between operators and clinics, leading to large variations in organ at risk (OAR) doses reported in the literature. The aim of this study was to develop an automated standardized VMAT-planning process for locoregional breast patients that could provide dose plans that were equal or better than 3D-CRT regarding target coverage and doses to organs at risk (OARs) for a large variation of patient cases.

Material/Methods:

CT-studies from 20 left sided breast patients in deep-inspiration breath hold were used to develop a script that automated all parts of the VMAT dose planning process. Planning was performed in Raystation 12A (RaySearch Laboratories AB, Stockholm, Sweden) with an Elekta Versa HD linac model (Elekta Solutions AB, Stockholm, Sweden). Delineation of target volumes and organs at risk were performed according to ESTRO-guidelines [1]. Dose prescription was a median dose of 40.05 Gy in 15 fractions to the CTV. For each case, a 3D-CRT plan with tangential beams matched with anterior-posterior beams cranially was developed for benchmarking. For VMAT plans, the automated scripted process initially evaluated and optimized the doses to the OARs (heart, lungs, contralateral breast, oeasophagus, thyroid gland, trachea, humeral head and the coronary artery) according to what was feasible for each individual patient thorough five optimization cycles. In a second stage, the weights of the minimum dose objectives for the target volumes were increased in a systematic way during five optimization cycles until the required target D98 values were obtained. Elimination of hot spots and improvement of target dose homogeneity was also included in this phase of the optimization. Complete automatic treatment plans were made in 30-45 min.