Analysis of the applicability of two‐dimensional detector arrays in terms of sampling rate and detector size to verify scanned intensity‐modulated proton therapy plans

Abstract

The introduction of advanced treatment techniques in proton therapy, such as intensitymodulated proton therapy, leads to an increased need for patient-specific quality assurance, especially an accurate treatment plan verification becomes inevitable. In this study, signal theoretical analysis of dose distributions in scanned proton therapy is performed to investigate the feasibility and limits of two-dimensional (2D) detector arrays for treatment plan verification. Methods: 2D detector arrays are characterized by two main aspects: the distance between the single detectors on the array or the sampling frequency; and the lateral response functions of a single detector. The analysis is based on single spots, reference fields and on measured and calculated dose distributions of typical intensity-modulated proton therapy treatment plans with and without range shifter. Measurements were performed with Gafchromic EBT3 films (Ashland Speciality Ingredients G.P., Bridgewater, NJ, USA), the MatriXX PT detector array (IBA Dosimetry, Schwarzenbruck, Germany) and the OCTAVIUS detector array 1500XDR (PTW-Freiburg, Germany) at an IBA Proteus PLUS proton therapy system (Ion Beam Applications, Louvain-la-Neuve, Belgium). Dose calculations were performed with the treatment planning system RayStation 6 or 8 (RaySearch Laboratories, Sweden). Results: The Fourier analysis of the data of the treatment planning system and film measurements show maximum frequencies of 0.06/mm for the plan with range shifter and 0.083/mm for the plan without range shifter. According to the Nyquist theorem, this corresponds to minimum required sampling distances of 8.3 and 6 mm, respectively. By comparison, the sampling distances of the arrays of 7.6 mm (MatriXX PT) and 7.1 mm (OD1500XDR) are sufficient to reconstruct the dose distributions adequately from measurements if range shifters are used, whereas some fields of the plans without range shifter violated the Nyquist requirement. The lateral dose response functions of the single detectors within the arrays have clearly higher frequencies than the treatment plans and thus the volume effect only slightly influences the measurements. Consequently, the array measurements show high gamma passing rates with at least 96 % and a good agreement between the investigated line profiles. Conclusion: The results indicate that the detector dimensions and sampling distances of the arrays are in most studied cases adequate not to substantially influence the measurement process when they are used for analyzing typical intensity-modulated proton therapy treatment plans. Nevertheless, clinical conditions have been identified, for instance treatment plans without range shifter, under which the Nyquist theorem is violated such that a full representation of the dose distributions with the measurements is not feasible. In these cases, analysis of measurements is limited to pointwise comparisons.

Investigating the lateral dose response functions of point detectors in proton beams

“…The film response was calibrated in a 50 mm diameter 150 MeV proton beam against the dose measured with a PTW Markus chamber type 23343 (0, 0.02, 0.05, 0.1, 0.21, 0.42, 0.84, 1.1, 1.7 and 4.2 Gy) at 2 cm depth in a polystyrene phantom. The calibration was performed following the procedure described in Brodbek et al (2020) but using two third-degree polynomial fit functions, one for the low and another for the high dose range, to improve the goodness of the fit over the entire dose range.…”

“…For general PBS with gaussian shaped pencil beams, previous studies indicate that the volume effect of detectors, which comprises the volume-averaging and density effects, does not or only slightly influence the measurement process (Brodbek et al, 2020;Furukawa et al, 2013;Moignier et al, 2017;Sahoo et al, 2010;Schwaab et al, 2011). In contrast, insight into the performance of point detectors in proton beams with reduced penumbrae is scarce.…”

“…The function K(x,y) acts as the convolution kernel transforming the undisturbed dose distribution D(x,y) into the measured signal distribution M(x,y) (Looe et al, 2013;Harder et al, 2014;Looe et al, 2015Looe et al, , 2015Poppinga et al, 2015;Looe et al, 2017;Brodbek et al, 2020):…”

“…Considering the shorter average secondary electron range in proton beams as compared to photons (Brodbek et al 2020) the influence of the density perturbation is anticipated to be smaller in proton beams, while the volume-averaging component remains the same.…”

The root cause analysis on failed patient‐specific measurements of pencil beam scanning protons using a 2D detection array with finite size ionization chambers

“…The commercially available 2D ionization array IBA MatriXX PT 9 , 10 is used for planar dose measurements. The MatriXX PT array consists of 1020 air‐vented plane‐parallel ionization chambers in a 32 × 32 matrix with a sampling distance of 7.62 mm over an area of 244 mm × 244 mm.…”