Table 3.
Physical phantoms developed by research group.
| Phantom (site) | Deformable/anthropomorphic | Motion reproducibility | Main features |
|---|---|---|---|
| WashU (any site) (Malinowski et al 2007) | No/no | • Target accuracy (mean ± SD) < 1 mm | • 3D axis and independent 1D vertical axis |
| • Motorized platform to carry phantom or dosimetry equipment | |||
| LuCa (lung) (Perrin et al 2017) | Yes (interior and exterior)/yes (high level of detail) | • Stable end in/exhale (<1 mm) | • Inflatable/deformable lungs, skeleton, muscles skin, solid heart, solid mobile tumour (can hold dosimetric films) |
| • Tumour position varied from day to day for a given intermediary in/exhale pressure | • Motion actuated by an air pump inflating the lungs | ||
| • MR-compatible with visible deforming lung features | |||
| Lung (Cheung and Sawant 2015) | Yes (interior and exterior)/yes (low level of internal detail) | • <2 mm day-to-day (ascribed to set-up) | • Deformable external shell |
| • Latex foam insert for lungs | |||
| • < 0.25 mm RMS intra-day | • Rigid foam diaphragm actuated by the WashU motion stage | ||
| Lung (Biederer et al 2006, Remmert et al 2007) | Yes (interior)/yes (animal heart and lungs, nodules, airways, no ribs) | • Maximal diaphragm displacement precision (SD) 1.90 mm (on CT), 1.47 mm (on MR) | • Porcine lung and heart explants with tracheal tube in saline solution, artificial pulmonary nodules |
| • Water-filled silicon diaphragm inflated or deflated by a water pump outside the MR room | |||
| • Reproducibility of intermediary phases not quantified | • MR-compatible | ||
| Lung (Serban et al 2008) | Yes (interior)/yes (only one lung with vasculature/airways features) | • Within image resolution (0.7 × 0.7 × 1.25 mm3) | • Lung (natural latex balloon filled with damp sponges) in water, thoracic cavity (Lucite), diaphragm (motor-actuated piston), tumour (Dermasol ellipsoid), vascular and bronchial bifurcation (nylon wires and Lucite beads) |
| Lung (Steidl et al 2012) | Yes (exterior)/yes (low level of internal detail, cubic tumour) | Target accuracy (mean ± SD) = 0 ± 0.09 mm (input versus log files) | • Artificial skeleton, rubber skin |
| • Tumour: PMMA cube with 20 slots for pinpoint ion chambers and 5 films | |||
| • Sternum-induced thoracic motion | |||
| • 6D robot-actuated tumour motion independent of thoracic motion | |||
| MAESTRO (lung) (Haas et al 2014) | Yes (ribcage only)/yes (no vasculature) | • Millimetre positioning precision | • Mechanically actuated ribs, stationary lungs, trachea and spine in hermetic skin (to be filled with water) |
| • Inter-cycle reproducibility <0.16 mm RMS | • Robot-actuated tumour motion | ||
| ELPHA (liver) (Ehrbar et al 2019) | Yes (interior)/yes (liver with vasculature) | • Reproducibility <0.32 mm RMS (inter- and intra-day) | • Soft silicon liver with vasculature (can hold dosimetric devices) |
| • Static inferior plate and motor-driven superior plate | |||
| • Ultrasound and CT contrast | |||