Evaluation of a novel anthropomorphic thorax phantom with a dynamic lung using 4DCT

Abstract

Purpose/Objective: Four-dimensional computed tomography (4DCT) has revolutionized the radiation therapy field by allowing visualization and tracking of movements in the target volume and organs at risk, leading to more accurate treatment procedures. However, the image quality and accuracy of quantitative information could be affected by severe artifacts arising from regular or irregular breathing and motion artifacts. These image quality measurements are mostly performed using phantoms, which do not represent entirely the complexities and motion of patients. Recently, the 3D-printing technology has been introduced into radiotherapy that allows for increased customization and generating anthropomorphic dynamic/deformable phantoms. This study aims to evaluate the 4DCT image quality and proof-of-concept of an in-house developed realistic dynamic/deformable phantom in terms of tumor trajectory and motion artifacts, which allows for more realistic motion evaluation than with more commonly used simplistic phantoms. Material/Methods: A novel anthropomorphic thorax phantom (Figure 1A) was manufactured with fused deposition modeling (FDM) printing by using tissue-equivalent materials that represented soft tissue, lung tissue and bone. These materials were selected based on the effective atomic number and relative electron density. The soft tissue structure was based on CT images of a real patient with, in addition, tissue-equivalent bone such as spine and ribs and a deformable and compressible lung, including bronchi and tumors with accurate mechanical properties for realistic compression. The compression was performed by a developed electronic lung compression system (LCS), that simulated realistic respiration induced breathing motion. This system was connected to a static chest movement system (CMS; Figure 1B). The latter system was used to connect the ANZAI belt and allowed to track the breathing phases by the 4DCT system. A snippet of the respiratory curve is shown in Figure 2A. Images were acquired in static 3DCT (reference volume) and 4DCT. The image acquisition was performed on a SOMATOM Definition Drive CT scanner (Siemens Healthineers) with a tube voltage of 120 kVp (Qr40). This scanner used a phase-based sorting algorithm to reconstruct at specific breathing phases (0% inhale, 25% inhale, 50% inhale, 75% inhale, 100% inhale, 75% exhale, 50% exhale, 25% exhale). The scan parameters for the 4DCT were chosen based on clinical practice and included a pitch of 0.14 s, a field-of-view of 500 mm and a CTDIvol of 22 mGy. In addition, reconstruction was performed with 3 mm slice thickness. In evaluation, two tumors that had different volumes(tumor 1 and 2; Figure 1A) were assessed by quantifying the center of mass and volume between the respiratory phases. In addition, the amplitude between the different tumors was evaluated to demonstrated the realistic motion induced by the LCS.