ESTRO 2024 - Abstract Book

S4081

Physics - Inter-fraction motion management and offline adaptive radiotherapy

ESTRO 2024

Purpose/Objective:

Molded thermoplastic masks are today the mainstream solution for patient immobilization during radiotherapy (RT). The emergence of 3D-printing brought new possibilities to producing patient-specific medical devices based on imaging data. The versatility of these technologies allows to rethink the design of immobilization devices and provides opportunities to overcome some of the disadvantages of thermoplastic masks. As the production of the devices does not involve the patient, they are spared from the discomfort during the molding process. An automated digital design and manufacturing can result in more consistent quality compared to the manual molding of thermoplastic masks, independent of the operator’s skills and the cooling process related variability. Unlike fully encompassing the body contour a more open device can be created, reducing anxiety for patients, targeting stable regions around the head potentially retaining high quality immobilization also in case of anatomical changes and treatment adaptation. In this abstract, a first iteration of an innovative design for an entirely 3D-printed double shell patient-specific head and neck immobilization device using CT imaging data is presented, including an extensive validation study for aspects of manufacturing, mechanical design and dosimetry. As a starting point, a skin-to-bone distance mapping was performed based on CT images of 6 patients to determine stable regions for immobilization of the head and neck region that potentially suffer less from anatomical changes during RT. The immobilization device was designed in-house based a CT image of a CIRS phantom model 731-HN using stable region information from patient data. The design considered patient comfort, usability of the device in practice and mechanical requirements including appropriate mechanical safety factors. Two devices of the same design were 3D-printed in PA12 using two different state-of-the-art 3D-printing technologies: Selective Laser Sintering (SLS) and Multi Jet Fusion (MJF). Both CT and MR imaging was acquired of the 3D-printed devices to assess compatibility with both these imaging modalities. For CT compatibility the average HU was determined and the presence of artefacts was assessed. For MR compatibility, both devices were imaged while submerged in water while accompanied by a plastic grid to assess artefacts and geometrical deformations caused by the devices. In-silico validation of the design was performed by assessing the deformation of the device under load using a Finite Element Method (FEM) static stress simulation. The simulation was verified by physically measuring the downward deflection using a dial gauge. High-resolution CT was used to assess the geometric accuracy and fit of the phantom in the printed devices. Geometric deviations of the digitized device with respect to the CAD model were determined by mapping the closest point distances. The positional reproducibility was assessed by acquiring 5 CT-scans while each time removing and reinstalling the phantom into the device. Translational and rotational deviations relative to the initial CT-scan and model were calculated using rigid registration. A dosimetric validation was performed by delivering static beams, two VMAT and one non-coplaner VMAT to the phantom placed in the MJF device. Point doses were measured inside the target volume using an A1SL IC (Standard Imaging, USA) and compared to dose calculations (Acuros, Varian MS, USA). Material/Methods:

Results: