A 4D-dosimeter and quality assurance phantom prototype was developed to quantify the effects of respiratory motion.

The dose distributions were measured using two-dimensional detectors that were mounted on a mobile platform capable of sinusoidal motion in one direction with different patterns using adjustable motion amplitude and frequency. The dose distributions were obtained from various treatment plans including conformal and intensity-modulated beams for both photon and proton therapy. Dose delivery and measurement were conducted using this 4D-dosimeter with the mobile phantom for different motion amplitudes (0–35 mm) and frequencies (0.25–0.33 Hz).

The increase in motion amplitude increased the blurring of the dose distributions at the beam edges along the direction of motion and led to large dose discrepancies. This produced larger dose deficits inside the treatment planning volume (PTV) and increasing dose deposition in the surrounding normal tissue with increasing motion amplitudes. For both the IMRT and VMAT-treatment plans, the dose profile for each increased amplitude increment showed a reproducible flattening of the penumbra at the beam edge, all changing around the 40–60% isodose line.

The 4D-dosimeter developed in this work provides a noble clinical tool to quantify the deviations in the dose distributions induced by respiratory motion.

The purpose of this study is to investigate quantitatively the correlation of displacement vector fields (DVFs) from different deformable image registration (DIR) algorithms to register images from helical computed tomography (HCT), axial computed tomography (ACT) and cone beam computed tomography (CBCT) with motion parameters.

CT images obtained from scanning of the mobile phantom were registered with the stationary CT images using four DIR algorithms from the DIRART software: Demons, Fast-Demons, Horn–Schunck and Lucas–Kanade. HCT, ACT and CBCT imaging techniques were used to image a mobile phantom, which included three targets with different sizes (small, medium and large) that were manufactured from a water-equivalent material and embedded in low-density foam to simulate lung lesions. The phantom was moved with controlled cyclic motion patterns where a range of motion amplitudes (0–20 mm) and frequencies (0·125–0·5 Hz) were used.

The DVF obtained from different algorithms correlated well with motion amplitudes applied on the mobile phantom for CBCT and HCT, where the maximal DVF increased linearly with the motion amplitudes of the mobile phantom. In ACT, the DVF correlated less with motion amplitudes where motion-induced strong image artefacts and the DIR algorithms were not able to deform the ACT image of the mobile targets to the stationary targets. Three DIR algorithms produce comparable values and patterns of the DVF for certain CT imaging modality. However, DVF from Fast-Demons deviated strongly from other algorithms at large motion amplitudes.

The local DVFs provide direct quantitative values for the actual internal tumour shifts that can be used to determine margins for the internal target volume that consider tumour motion during treatment planning. Furthermore, the DVF distributions can be used to extract motion parameters such as motion amplitude that can be extracted from the maximal or minimal DVF calculated by the different DIR algorithms and used in the management of the patient motion.