A 2D strip ionization chamber array with high spatiotemporal resolution for proton pencil beam scanning FLASH radiotherapy

Abstract

Purpose Experimental measurements of two‐dimensional (2D) dose rate distributions in proton pencil beam scanning (PBS) FLASH radiation therapy (RT) are currently lacking. In this study, we characterize a newly designed 2D strip‐segmented ionization chamber array (SICA) with high spatial and temporal resolution and demonstrate its applications in a modern proton PBS delivery system at both conventional and ultrahigh dose rates. Methods A dedicated research beamline of the Varian ProBeam system was employed to deliver a 250‐MeV proton PBS beam with nozzle currents up to 215 nA. In the research and clinical beamlines, the spatial, temporal, and dosimetric performances of the SICA were characterized and compared with measurements using parallel‐plate ion chambers (IBA PPC05 and PTW Advanced Markus chamber), a 2D scintillator camera (IBA Lynx), Gafchromic films (EBT‐XD), and a Faraday cup. A novel reconstruction approach was proposed to enable the measurement of 2D dose and dose rate distributions using such a strip‐type detector. Results The SICA demonstrated a position accuracy of 0.12 ± 0.02 mm at a 20‐kHz sampling rate (50 μs per event) and a linearity of R2 > 0.99 for both dose and dose rate with nozzle beam currents ranging from 1 to 215 nA. The 2D dose comparison to the film measurement resulted in a gamma passing rate of 99.8% (2 mm/2%). A measurement‐based proton PBS 2D FLASH dose rate distribution was compared to simulation results and showed a gamma passing rate of 97.3% (2 mm/2%). Conclusions The newly designed SICA demonstrated excellent spatial, temporal, and dosimetric performances and is well suited for commissioning, quality assurance, and a wide range of clinical applications in proton PBS clinical and FLASH‐RT.

Characterization of 250 MeV Protons from the Varian ProBeam PBS System for FLASH Radiation Therapy

“…This is because the Lynx detector's count range, 0 to 1000, limits its ability to measure low-dose tails with reasonable accuracy; ± 1.5 cm is coincident with counts falling below 10 −3 (normalized value). This is where the profile deviates significantly from a single Gaussian significantly, as shown previously by several investigators [ 29 , 31 , 32 ]. R-square and root-mean-square-error values are obtained by considering the points above the 10 −3 threshold.…”

“…Although they have not reached FLASH DRs, their study shows that beam spot size and energy are not DR-dependent. However, under FLASH conditions, beam optics (due to the space charge effect of densely bunched protons) and beamline configuration (like removal of some component or change of steering parameters) can change, prompting spot profile comparison of conventional and FLASH beams [ 28 , 29 ]. This type of comparison was not performed in this work and is a limitation of this study.…”

Time‐resolved dose rate measurements in pencil beam scanning proton FLASH therapy with a fiber‐coupled scintillator detector system

“…In another study, the measured dose profile found by a 2D strip segmented ionization chamber array was used for PBS dose rate calculations. 28 A functional form of the lateral dose profile that includes the tails of the dose profile was used in that study and the PBS dose rate pattern yielded comparable results with our PBS dose rate map using the measured dose profile (Figure 8b).…”

“…The PBS dose rate map calculated previously for the same field and beam parameters 6 yielded similar results as the case of a Gaussian profile in this study (max DR PBS95% = 531 Gy/s, Figure 8c), but the maximum PBS dose rate (418 Gy/s) was lower due to a larger Gaussian width (σ x = 3.9 mm and σ y = 4.1 mm) and lower field dose rate (77 Gy/s) used in the simulation. In another study, the measured dose profile found by a 2D strip segmented ionization chamber array was used for PBS dose rate calculations 28 . A functional form of the lateral dose profile that includes the tails of the dose profile was used in that study and the PBS dose rate pattern yielded comparable results with our PBS dose rate map using the measured dose profile (Figure 8b).…”

“…The deviation from a Gaussian spot dose profile has been observed in previous studies where the dose profile in water was modelled using an additional Gaussian function 27 or a combination with two Gaussian functions and a Lorentzian function. 18,28 In this study, two Gaussian functions and a Lorentzian function were used since it was more accurate in describing the dose profile in a large range of distances. It is important to note that the spot dose profile characterization requires proper calibration and cannot be performed on the raw detector signal due to the non-linear dose rate dependency.…”