Evaluation of radiological properties and anisotropy with air channels analysis in 3D-printed flexible lung-mimicking materials for radiotherapy

Abstract

Objective. Three-dimensional (3D) printing is increasingly used for radiotherapy quality assurance (QA) phantoms, yet, reproducing the structural heterogeneity and radiological properties of lung tissue remains challenging. This study evaluates thermoplastic polyurethane (TPU) for dynamic lung-mimicking phantoms, selected for its flexibility while preserving print-defined geometry, with a focus on radiological equivalence, directional anisotropy, and the detectability of sub-millimetre air channels. Approach. Eleven TPU materials with various colours and Shore hardness (63-82) were printed into gyroid-patterned inserts of varying infill densities. Effective atomic number (Zeff) and relative electron density (RED) were determined using dual-energy computed tomography (CT). Anisotropy and internal air channels were assessed in five orientations using micro-CT, clinical CT, and flat-panel detector (FPD) imaging, and compared to a voxelised digital model derived from G-code toolpaths. Main results. Measured Zeff ranged from 6.3 +/- 0.6 to 11.1 +/- 0.2, with pigment-driven variation up to 66% within identical material categories. Seven materials achieved lung mimicking Zeff (around 7.55). RED increased with infill and mimicked lung references (0.28-0.43) at moderate infill. The low-infill insert contained 0.7 mm air channels visible in all modalities. The high-infill insert contained 0.3 mm channels, resolved accurately by the micro-CT and FPD but not reliably resolved by clinical CT due to resolution limitations. The digital model indicated diagonal anisotropy, micro-CT and FPD indicated near-isotropy in low and high infill, while CT showed apparent anisotropy due to its resolution limitations. Significance. TPU-based gyroid phantoms are suitable lung-mimicking candidates for radiotherapy QA of imaging and dosimetry. Their periodic air-channels are reliably resolved by high-resolution imaging (micro-CT or FPD), but may be distorted by clinical CT, particularly in resolution-limited orientations, the digital model supports pre-print evaluation of these air-channels. Because undetected internal heterogeneities may affect dose calculation accuracy, high-resolution imaging when available, is preferred for assessing the internal structure of 3D-printed phantoms.