ESTRO 2024 - Abstract Book

S4017

Physics - Inter-fraction motion management and offline adaptive radiotherapy

ESTRO 2024

Radiotherapy in head and neck cancer (HNC) requires the use of immobilization masks to ensure precise repositioning and minimize movement. Furthermore, MR scans are superior compared to CT for delineating HNC. The conventional 5-point HNC positioning mask has to be used in combination with single-loop receive coils which results in relatively low-quality MR scans. To address this limitation, we investigated a double shell HNC positioning system, which allows the use of a phased-array coil, thereby improving image quality. Our objective is to assess the effectiveness of the double shell HNC positioning system in reducing inter- and intrafraction displacements, aiming for repositioning accuracy and motion reduction comparable to the conventional 5-point HNC positioning mask.

Material/Methods:

Fig.1: Regions of interest for motion analysis defined by clip boxes: (a, blue) skull, (b, light orange) mandible, (c, orange) vertebrae C4-C6, (d, red) shoulder, (e, dark blue) head-and-neck area, and (f, light blue) oropharynx region (specific to volunteer analysis). The masks examined were the Posicast 5-point mask (Posicast®) from Civco Medical Solutions (Orange City, Iowa, USA), currently used at our department, and the Double Shell Positioning System (DSPS) Prominent® developed by MacroMedics BV (Moordrecht, The Netherlands). As a preclinical test, we acquired MR scans of volunteers (n=12) and patients (n=5) wearing a personalized DSPS at different time points. Participants underwent two imaging sessions on separate days. Interfraction displacements between days were simulated by removing the mask and repositioning the participant. Interfraction displacements between weeks were measured between the two scanning sessions. To investigate intrafraction displacements, scans were acquired after 5, 10, and 30 minutes of scan time. To compare interfraction displacements in the DSPS and Posicast, CBCTs acquired for daily set-up verification and planning CTs from patients treated in Posicast (n=50) masks were collected. For each scan set and for several region of interests, a rigid registration with a clipbox around the region of interest (Fig. 1) was performed following the methodology of Houweling et al. [1]. For CBCTs and planning CTs acquired with Posicast masks, the planning CTs were considered the fixed scans. For MRIs obtained in DSPS masks, the first MR scan was chosen as the fixed scan.

Using the displacements in the x-, y- and z-direction resulting from the rigid registration, the absolute displacement was calculated and compared.

Results: