Evaluating the impact of volume averaging in pencil beam scanning proton therapy

Introduction

Background

Proton therapy, particularly with pencil beam scanning (PBS), stands as a cutting-edge treatment for cancer, offering unparalleled precision in delivering therapeutic doses to tumors while minimizing damage to healthy tissues (1). Unlike traditional passive scattering proton therapy, a PBS proton system utilizes magnetically steered, spot-by-spot delivery (2,3). The Gaussian-shaped profiles of PBS spots feature forward peaks, adequately resembling the profile characteristics of flattening filter free (FFF) X-ray beams, which are known to exhibit volume averaging in dose measurement (4).

Volume averaging is well characterized in photon therapy. For FFF X-rays, it results from the non-uniform radial dose distribution integrated over the collecting area of an ion chamber (4). Small-volume chambers are therefore required to reduce this effect, which can cause output measurement errors (5-7). Additionally, established methods, such as applying correction factors (4,8,9), have proven effective in mitigating the volume averaging effect in photon therapy.

Rationale and knowledge gap

In contrast to FFF X-rays, the PBS proton system delivers dose spot by spot, and the composite profile varies with spot spacing and spot size (10). This variability presents a challenge for parallel-plate (PP) ion chambers, which are recommended for proton reference dosimetry (11), in measuring dose without volume averaging. In proton therapy, both measurement and simulation studies have reported volume averaging in output dose measurements with the PP chambers (12,13).

This challenge is particularly relevant to beam configuration during machine commissioning, a process of critical clinical importance. First, the dose distributions are measured in a phantom and are used as input parameters for the treatment planning system (TPS), which in turn calculates dose distributions that are verified against measurements. However, as noted earlier for proton systems, the output dose measured with a PP chamber exhibits volume averaging. This raises an important question whether TPS verification should rely on point doses or volume-averaged doses that better reflect the measurement process.

Objective

Our study aims to address this question by quantifying the difference between the two approaches. We demonstrate volume averaging in TPS by comparing the central-axis point dose and a volumetric dose contoured within the chamber. Our goal is to develop an optimal strategy that supports accurate dose measurement and safe patient treatment in PBS proton therapy.

Methods

TPS simulation

Proton energies of 70, 150 and 225 MeV were selected to span the clinical range of therapeutic beams available on the ProteusPlus PBS therapy machine (IBA, Lovain-La-Neuve, Belgium). A PPC05 PP chamber (IBA Dosimetry, Schwarzenbruck, Germany) was computed tomography (CT)-scanned and incorporated in a virtual water phantom created in RayStation version 11A TPS (RaySearch Laboratories, Stockholm, Sweden). For each plan, an anterior-posterior (AP) beam was directed perpendicularly onto the chamber surface along the central axis. A forward planning technique was used. No range shifter or snout was included. The spot pattern was rectangular, with a 0.1 cm grid—the smallest available in the TPS. Dose calculations were performed using the Monte Carlo (MC) algorithm.

The experiment was designed following the guidelines outlined in the RayStation 11A TPS manual for beam commissioning data specification, consistent with our clinical commissioning procedures. The manual recommends performing dose acquisitions at a depth approximately midway between 1 cm and one half of the position of the Bragg peak maximum (10). This is in alignment with protocols of proton reference dosimetry, which have shown that measured dose can be misrepresented in high-gradient areas such as the Bragg peak or distal falloff (14-16). Based on this, the buildup depth was defined as the distance from the isocenter to the chamber surface: 1.47 cm for 70 MeV, 4.37 cm for 150 MeV, and 8.35 cm for 225 MeV. A range of fixed spot spacings, from 0.18 cm to 0.6 cm, was selected to represent different degrees of spot overlap across the chamber’s detecting area, including the 0.25 cm spot spacing recommended for commissioning. To ensure consistency with the commissioning values, 0.3 monitor units (MU) per spot were used.

The doses were obtained in the TPS using two methods: point-based and volumetric-based. In accordance with the IAEA TRS-398 guideline for proton reference dosimetry (11), a dose specification point (DSP) was placed along the central axis of the AP beam on the inner surface of the entrance window, from which the point dose (Dptps) was acquired. For the volumetric dose (Dvtps), values were obtained from the contoured active measurement element of the PPC05 chamber, based on its schematic design (Figure 1). The cylindrical active collecting element, with a height of 0.06 cm and a diameter of 0.99 cm, resulted in a volume of 0.046 cm³.

Figure 1 Apeny PPC05 parallel-plate ion chamber. (A) Schematic design showing the entrance window in a yellow rectangle, and the active measurement element in a red rectangle. (B) CT scan displaying the active measurement element and the DSP location (red dot) along the central the central axis. CT, computed tomography; DSP, dose specification point.

To quantify the relationship between the active measuring area and the spot spacing, the concept of dose density (ρdose) was introduced:

ρdose=Nspot×M1×M2×A−1[1]

where ρdose, is the measured dose per unit area of the chamber’s active measurement element (cGy·cm⁻²). Nspot is the total number of spots, M₁ is the dose per MU, and M₂ is the MU per spot. A is the unit area (cm²). Dose density for each plan was calculated using Eq. [1].

Phantom verification

Plans with varying dose densities, ρdose , were delivered for the three selected energies—70, 150 and 225 MeV—using an IBA Blue Water Phantom2 (IBA Dosimetry GmbH, Schwarzenbruck, Germany) and the ProteusPlus PBS therapy machine. Following the IAEA TRS-398 guideline (11), measured doses (D^m) were calculated from raw data acquired using the PPC05 PP chamber, which is suitable for PBS reference dosimetry, as supported by recent work (17). The D^m values were then compared with the TPS-predicted values of Dptps and Dvtps, as shown in Figure 2.

Figure 2 Comparison of in-phantom and in-TPS measured doses for (A) 70 MeV, (B) 150 MeV, and (C) 225 MeV. The doses acquired in TPS are either point-based (Dptps) or volumetric-based (Dvtps). The percentage of difference is calculated as ΔPhantom=(Dporvtps−DmDm)×100%. TPS, treatment planning system.

Volume averaging for composite spot profiles were evaluated, as depicted in Figure 3. According to RayStation TPS definitions, the spot size represents the standard deviation (σ) of the Gaussian-shaped profile defined in air at the isocenter, and the spot spacing refers to the peak-to-peak distance between two adjacent spots (10). Composite profiles were plotted for spot spacings of 0.6, 0.25, and 0.18 cm, corresponding to dose densities of 0.1 cGy·cm⁻², 0.4 cGy·cm⁻², and 0.8 cGy·cm⁻², respectively. Each composite profile consists of five superposed spots, with σ values of 0.629, 0.383, and 0.287 cm for 70, 150, and 225 MeV, respectively, as determined from RayStation TPS beam configuration.

Figure 3 Volume averaging in composite spot profiles. The superposed dose profile (green) is formed by summing individual PBS spots. (A-C) Dose densities of 0.1, 0.4, and 0.8 cGy/cm², respectively. The blue strip represents the ROI for the DSP, and the grey strip represents the ROI of the PPC05 active measurement volume. The area difference between the DSP and chamber ROI under the composite profile is denoted as ∆_Area. DSP, dose specification point; ROI, region of interest.

Statistical analysis

TPS-calculated doses were deterministic outputs and were not subjected to statistical analysis. For phantom measurements, three repeated readings were acquired at each measurement point, and the mean values were used for comparison with the TPS-calculated doses.

Ethical considerations

This study did not involve human participants, identifiable personal data, or animal subjects. All analyses were performed using TPS calculations and phantom-based measurements.

Results

Significant discrepancies between point-based and volumetric-based dose acquisition methods were observed. The in-TPS results are presented in the “TPS measurement” section and the in-phantom results in the “Phantom measurement” section. Additionally, the “Volume averaging” section estimates the impact of volume averaging. The discrepancies between the two methods were small when adhering to RayStation TPS’s recommended commissioning settings: performing dose measurement at a depth approximately midway between 1 cm and one half of the position of the Bragg peak maximum, with a spot spacing of 0.25 cm and a 10×10 cm² field (10).

TPS measurement

The deviation between Dptps and Dvtps was highest at the lowest dose density, ρdose of 0.1 cGy·cm⁻² for all energies. The discrepancy reached up to 4.5% for 70 MeV, 2.2% for 150 MeV, and 4.3% for 225 MeV, as shown in Table 1. However, it decreased to below ±1% when using the TPS recommended commissioning settings (ρdose of 0.4 cGy·cm⁻²), and remained below ±1% at higher dose densities. Under the same commissioning conditions, the differences between the isocenter doses and Dptps were −13.2% (70 MeV), −11.3% (150 MeV), and −9.4% (225 MeV).

Phantom measurement

Figure 2 summarizes the in-phantom measurement results. The deviation between the in-phantom and in-TPS measurements was quantified as ΔPhantom=(Dporvtps−DmDm). The maximum ΔPhantom occurred at the minimum ρdose of 0.1 cGy·cm⁻² for all energies. Specifically, ΔPhantom exceeded ±5% for volumetric-based measurement at 70 MeV, whereas point-based measurements remained below ±2% for all energies. At commissioning conditions recommended, ρdose of 0.4 cGy·cm⁻², the ΔPhantom values were below ±1%.

Volume averaging

Figure 3 illustrates the results of volume averaging in composite spot profiles. The differences between the two shaded areas quantify the discrepancies associated with volume averaging. The largest area difference, indicating the maximum volume averaging, occurred at the lowest ρdose of 0.1 cGy·cm⁻² across all energies. The area discrepancies decreased as ρdose increased.

Discussion

Key findings

Our study demonstrated the clinical impact of volume averaging in PBS proton therapy. We further revealed how variations in spot size and spacing contribute to this effect, with discrepancies between point and volume doses more pronounced at lower dose densities (<0.2 cGy·cm⁻²). Point dose acquisition in TPS also showed better agreement with in-phantom measurement, supporting its use over volume dose for minimizing the volume averaging. For accurate TPS configuration, our findings also highlight the importance of adhering to vendor-recommended commissioning reference conditions.

Strengths and limitations

Prior studies on volume averaging have focused on in-phantom dose measurements (4,8,9,14,15). Our work extends this discussion to in-TPS dose acquisition for PBS systems—an overlooked but critical topic because of its direct relevance to beam-configuration choices during commissioning baselining, which subsequently propagate into all downstream quality assurance (QA) and treatment planning activities. Using the RayStation TPS, one of the most widely adopted systems in proton therapy, we provide actionable guidance to minimize the clinical impact of TPS-based volume averaging.

As demonstrated in Figure 3, we reveal the underlying mechanism of volume averaging. To the best of our knowledge, this mechanism has not been previously described. Our workflow for optimizing in-TPS dose acquisition conditions begins with the point- versus volumetric-dose comparison in Table 1, which identifies dose densities of 0.4 cGy/cm² and above as appropriate reference settings. Subsequent in-phantom validation confirms this finding and further shows that discrepancies become more pronounced at a higher dose density of 0.8 cGy/cm². The concepts introduced here—particularly dose-density characterization and the selection of reference conditions—are generalizable to beam configuration in other proton TPS platforms.

Importantly, the described mechanism is not dependent on chamber type. Although our analysis used the PPC05 ionization chamber, volume-averaging behavior is governed by the ratio between detector cavity size and DSP. Because commonly used reference chambers (e.g., PTW 30013, NE-2571, Advanced Markus, and Exradin A12) all have active diameters much larger than the DSP, the qualitative patterns shown in Figure 3 also extend to these detectors. This broadens the relevance of our findings beyond a single chamber model.

This study has limitations. First, the physical thickness of the PPC05 entrance window (0.1 cm), rather than its water equivalent thickness (WET) of 0.156 cm, was used for dose calculations. The WET of the chamber’s C-522 entrance window was calculated by multiplying its mass density with the relative stopping power ratio of C-522 to water (16,18,19). A quantitative check showed that the resulting differences in Dptps and Dvtps were within ±1% at 70 MeV and became even smaller at 150 and 225 MeV due to the shallow dose gradient at the selected depths. Therefore, the impact on volume-averaging ratio is negligible. Similarly, Figure 3 did not account for variations in stopping power, which would act as a uniform scaling factor on both Dptps and Dvtps and therefore leaving the volume-averaging ratio unchanged. Third, the in-air spot size (σ) was used for spot superposition in Figure 3, even though the in-water σ at the chamber measurement plane is slightly larger, with a maximum estimated increase of 0.05 cm. This difference has a negligible impact on the superposition concept, as demonstrated in Figure S1.

Comparison with similar research

Existing proton dosimetry studies have focused primarily on chamber-dependent dose measurements in phantom. American Association of Physicists in Medicine (AAPM) Task Group (TG) 224 and TG 185 both recommend small-volume ion chambers to mitigate volume averaging in regions with steep dose gradients, such as the Bragg peak, modulation edges, or narrow PBS spots (14,15). Recent proton FLASH studies have similarly quantified chamber-size-dependent under-response in highly nonuniform fields (12,13).

In contrast to these in phantom investigations, our work develops a framework to minimize volume averaging in TPS dose acquisition, where the effect is not governed by chamber choice. In Figure 3, we demonstrate that the governing factor is the ratio between DSP and chamber cavity size. Because this ratio is small for virtually all commercial ion chambers, the resulting volume-averaging behavior in TPS is largely independent of chamber model. This extends the scope of our work into PBS beam configuration, where appropriate TPS dose-acquisition conditions are key to limiting volume averaging effect.

Explanations of findings

Despite the composite dose profile in Figure 3 resembling a non-uniform FFF beam in photon therapy, it is noteworthy that volume averaging differs between photon and PBS proton therapy due to their distinct beam characteristics. In photon therapy, beam profiles remain consistent once field size and measurement depth are fixed. As a result, volume averaging primarily depends on detector geometry and can be mitigated with correction factors (20,21). In contrast, in PBS proton therapy, the composite beam profile varies with spot size and spacing, creating an intrinsic, spatially dependent volume averaging that is not easily resolved. This variation is illustrated in Figure 3 and further supported by phantom measurements in the “Phantom measurement” section.

Despite differing root causes, mitigating volume averaging requires minimizing detector size - an approach proven effective in FFF X-rays, where smaller detectors reduce perturbation from non-uniform radial profile integrated over the detecting area (14,15). This principle also applies to PBS proton, where the smallest available ‘detector’ in TPS, in this context the DSP—a voxel point—should be used. In our study, the DSP possesses a finite volume equivalent to a voxel size of (0.098, 0.098, 0.0625 cm), derived from the smallest slice thickness of 0.0625 cm [50 cm display field of view (DFOV)] used for CT scanning, and the Dptps represents the dose at the center of the voxel. In contrast, the PPC05 has a much larger cylindrical region of interest (ROI) with a diameter of 0.99 cm and a height of 0.06 cm, making the ROI of the DSP approximately ten times smaller. The size difference accounts for the improved agreement between Dptps and D^m, suggesting reduced volume averaging with point-based acquisition.

During beam configuration, chamber measurements—volumetric by nature—are entered into the TPS as point dose inputs (15,22), inherently discarding spatial dose distribution information. Consequently, volumetric dose acquisition in TPS not only fails to mitigate volume averaging but also introduces systematic errors. While negligible in photon therapy, the discrepancy becomes significant in PBS proton therapy due to the intrinsic volume averaging from discrete spot delivery.

Implications and actions needed

We observed that the commissioning reference conditions recommended by RayStation TPS, with a dose density of 0.4 cGy/cm², yielded deviations below ±1% from phantom verification. Therefore, dose acquisition under the reference conditions—at a depth midway between 1 cm and half the Bragg peak maximum range—is critical for proper TPS configuration, which serves as the foundation for accurate dose calculation, reliable QA baselines (14), and ultimately, safer patient treatments. Our results show that using volumetric dose values or deviating from recommended reference conditions can introduce baseline errors up to 4.5% (Table 1), which propagate into downstream QA, treatment planning and dose delivery. To minimize baseline uncertainties, point-dose acquisition should be used during beam commissioning.

Conclusions

The PBS proton systems exhibit intrinsic volume averaging due to the superposition of spot profiles. We emphasize the importance of employing point-based rather than volumetric-based dose acquisitions in TPS to minimize volume averaging, which is the root cause of the significant discrepancies observed between the two methods at specific combinations of spot size and spot spacing. This further underscores the necessity of adhering to the recommended settings in TPS during beam commissioning, ensuring accurate and safe treatment delivery to patients.

Acknowledgments

None

Footnote

Data Sharing Statement: Available at https://tro.amegroups.com/article/view/10.21037/tro-25-40/dss

Peer Review File: Available at https://tro.amegroups.com/article/view/10.21037/tro-25-40/prf

Funding: None.

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://tro.amegroups.com/article/view/10.21037/tro-25-40/coif). The authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. This study did not involve human participants, identifiable personal data, or animal subjects. All analyses were performed using treatment planning system calculations and phantom-based measurements.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Gillin MT, Sahoo N, Bues M, et al. Commissioning of the discrete spot scanning proton beam delivery system at the University of Texas M.D. Anderson Cancer Center, Proton Therapy Center, Houston. Med Phys 2010;37:154-63. [Crossref] [PubMed]
2. Hopfensperger KM, Li X, Paxton A, et al. Measurements of fetal dose with Mevion S250i proton therapy system with HYPERSCAN. J Appl Clin Med Phys 2023;24:e13957. [Crossref] [PubMed]
3. Rahimi R, Taylor M, Li X, et al. Fetal dose assessment in a pregnant patient with brain tumor: A comparative study of proton PBS and 3DCRT/VMAT radiation therapy techniques. J Appl Clin Med Phys 2024;25:e14394. [Crossref] [PubMed]
4. Sudhyadhom A, Kirby N, Faddegon B, et al. Technical Note: Preferred dosimeter size and associated correction factors in commissioning high dose per pulse, flattening filter free x-ray beams. Med Phys 2016;43:1507-13. [Crossref] [PubMed]
5. Benedict SH, Yenice KM, Followill D, et al. Stereotactic body radiation therapy: the report of AAPM Task Group 101. Med Phys 2010;37:4078-101. [Crossref] [PubMed]
6. Buchegger N, Grogan G, Hug B, et al. CyberKnife reference dosimetry: An assessment of the impact of evolving recommendations on correction factors and measured dose. Med Phys 2020;47:3573-85. [Crossref] [PubMed]
7. Šegedin N, Hršak H, Babić SD, et al. Determination of volume averaging correction factors using an elliptical absorbed dose model for Gamma Knife Perfexion. J Appl Clin Med Phys 2023;24:e14109. [Crossref] [PubMed]
8. McEwen M, DeWerd L, Ibbott G, et al. Addendum to the AAPM's TG-51 protocol for clinical reference dosimetry of high-energy photon beams. Med Phys 2014;41:041501. [Crossref] [PubMed]
9. Das IJ, Francescon P, Moran JM, et al. Report of AAPM Task Group 155: Megavoltage photon beam dosimetry in small fields and non-equilibrium conditions. Med Phys 2021;48:e886-921. [Crossref] [PubMed]
10. RayStation 11A Beam Commissioning Data Specification. Available online: https://www.raysearchlabs.com
11. International Atomic Energy Agency. Absorbed Dose Determination in External Beam Radiotherapy, Technical Reports Series No. 398. IAEA, Vienna; 2000.
12. Darafsheh A, Bey A. Implementation of a proton FLASH platform for pre-clinical studies using a gantry-mounted synchrocyclotron. Phys Med Biol 2025;70:105008. [Crossref] [PubMed]
13. Darafsheh A, Hao Y, Zhao X, et al. Spread-out Bragg peak proton FLASH irradiation using a clinical synchrocyclotron: Proof of concept and ion chamber characterization. Med Phys 2021;48:4472-84. [Crossref] [PubMed]
14. Arjomandy B, Taylor P, Ainsley C, et al. AAPM task group 224: Comprehensive proton therapy machine quality assurance. Med Phys 2019;46:e678-705. [Crossref] [PubMed]
15. Farr JB, Moyers MF, Allgower CE, et al. Clinical commissioning of intensity-modulated proton therapy systems: Report of AAPM Task Group 185. Med Phys 2021;48:e1-e30. [Crossref] [PubMed]
16. ICRU Report 78, Prescribing, Recording, and Reporting Proton-Beam Therapy. International Commission on Radiation Units and Measurements; 2007. Available online: https://www.icru.org/report/prescribing-recording-and-reporting-proton-beam-therapy-icru-report-78
17. Lee E, Lourenço AM, Speth J, et al. Ultrahigh dose rate pencil beam scanning proton dosimetry using ion chambers and a calorimeter in support of first in-human FLASH clinical trial. Med Phys 2022;49:6171-82. [Crossref] [PubMed]
18. Palmans H, Medin J, Trnková P, et al. Gradient corrections for reference dosimetry using Farmer-type ionization chambers in single-layer scanned proton fields. Med Phys 2020;47:6531-9. [Crossref] [PubMed]
19. NIST PSTAR Database. Available online: https://physics.nist.gov/PhysRefData/Star/Text/PSTAR.html
20. Li X, Oare CC, Rahimi R, et al. A Feasible Approach for Assessing the Collecting Capacity of a Parallel-Plate Ionization Chamber for Accurate Dose Acquisition in Eclipse Treatment Planning System. In: AAPM 65th Annual Meeting & Exhibition, AAPM; 2023.
21. Das IJ, Cheng CW, Watts RJ, et al. Accelerator beam data commissioning equipment and procedures: report of the TG-106 of the Therapy Physics Committee of the AAPM. Med Phys 2008;35:4186-215. [Crossref] [PubMed]
22. Geurts MW, Jacqmin DJ, Jones LE, et al. AAPM MEDICAL PHYSICS PRACTICE GUIDELINE 5.b: Commissioning and QA of treatment planning dose calculations-Megavoltage photon and electron beams. J Appl Clin Med Phys 2022;23:e13641. [Crossref] [PubMed]