Session Item

For reference dosimetry in MRgRT, k_B,Q is used to correct for the impact of the magnetic field on the chamber calibration coefficient. It has been demonstrated that for accurate simulation of k_B,Q the dead volume (DV) must be considered. This work goes one step further by analysing the different contribution of all chamber components to k_B,Q.

Based on the blueprints the Farmer-type chamber PTW 30013 geometry was modelled in PENELOPE (2014). Two DVs were modelled; one based on published data (Pojtinger et al, 2019) (DV₁) and one determined using FEM simulations (DV₂). Radiation fields (10x10 cm²) were based on phase space files of a ⁶⁰Co source and a 7MV MRI-linac. The model was validated against measurements with a setup using an electromagnet and a ⁶⁰Co source (0 – 1.5 T) and an MRI-linac (0 and 1.5 T) (de Prez et al, 2019). Simulations were performed with the PENMAIN routine with magnetic field routines implemented for both parallel and perpendicular orientations of the chamber and the magnetic (B) field, and for several B-field strengths between 0 T and ±1.5 T. To study the dose contribution to the sensitive volume (SV = cavity – DV) from the electrons produced in certain components of the chamber (wall, central electrode (CE), guard, stem, phantom), the labelling of the particles was implemented in PENMAIN. A separate model with each solid component of the chamber modelled as liquid water (indicated by PTW 30013 (water)) was used to investigate the impact of material choice on k_B,Q.

The simulated k_B,Q results agree better with the measured k_B,Q when the DV is considered. The difference between measured k_B,Q and simulated k_B,Q reduces from at maximum 0.6% for DV₁ to 0.4% for DV₂ at B = 0.75 T in ⁶⁰Co. This demonstrates the accuracy of the model, and it shows that small changes in the DV may have a small impact in the k_B,Q.

Figures 1 and 2 show the relative contribution of each component to the effect of k_B,Q (i.e., k_B,Q-1). The negative value in the figure means an increase in the contribution from that component when the B is present.

Figure 1 demonstrates that the contribution to the SV dose from the bodies closer to the SV is higher than without B. For parallel orientation, the guard and the stem have higher impact on the dose to the SV. For the perpendicular orientation the wall contribution is more important. For all orientations the CE has a similar increased contribution

Figure 2 shows that the general behaviour of the contribution to the dose of the different components is similar. Overall, the impact to the dose in SV is reduced when the material of each component is modelled as liquid water.

Small components of the chamber impact k_B,Q considerably. Therefore, chamber design and, to a lesser extent, choice of material affects k_B,Q, and an accurate model of the chamber components and its further validation are important for correct calculations of k_B,Q.