Abstract
Aim:
The aim of the study was to investigate the dose calculation accuracy of volumetric modulated arc therapy (VMAT) control point (CP) discretisation approaches in Pinnacle and RayStation treatment planning systems (TPSs).
Methods:
The 6MV beam model for Elekta linear accelerator with Agility head generated in Pinnacle and RayStation was used in this study. The impact of CP discretization on dose calculation was assessed using 10 prostate and ten breast VMAT plans generated in each TPS. In Pinnacle, VMAT plans were generated using both 4° and 2° CP gantry spacing, while RayStation plans used 2° spacing. To assess the dosimetric differences due to CP discretization, VMAT plans from Pinnacle were recalculated in RayStation, and vice versa. The impact of CP discretization on target volume and critical structures was assessed by comparing dose volume histogram metrics. The agreement between ArcCheck-measured and TPS-calculated doses was assessed using gamma analysis with a 2%2 mm tolerance criterion.
Results:
The prostate and breast plans with 2° spacing generated in Pinnacle and recalculated in RayStation showed a mean (Std Dev) target dose agreement of 1.0 (0.4)% and 0.3 (0.8)%, respectively. The similar comparison for RayStation-generated plans recalculated in Pinnacle was 0.8 (0.2)% and 2.4 (0.2)%, respectively. The ArcCheck measurements of Pinnacle-generated prostate and breast plans showed a mean gamma pass rate of 96.5 (1.2)% and 91.5 (5.2)%, respectively. Similarly, RayStation-generated prostate and breast plans had a mean gamma pass rate of 96.7 (2.3)% and 95.4 (3.7)%, respectively.
Conclusions:
The CP discretization approach implemented in RayStation results in improved dose calculation accuracy of VMAT plans, particularly for plans with larger CP spacing.
Keywords: Control point, discretization, dose calculation, treatment planning system, volumetric modulated arc therapy
INTRODUCTION
Volumetric modulated arc therapy (VMAT) is a radiotherapy treatment technique which delivers a dose to a patient utilizing rotation or a treatment arc. The planning and delivery of VMAT beams is highly complex, with this technique able to vary the gantry rotation speed, the dose rate and the multi-leaf collimator (MLC) position and speed.[1,2] As such, inverse treatment planning has been utilized in treatment planning for this technique, with the treatment planning system (TPS) provided with objectives for the treatment plan, and the TPS finding the most optimal solution.[3]
Inverse treatment planning was introduced initially for intensity modulated radiation therapy, which utilized multiple MLC apertures, or control points (CPs), at static gantry angles. VMAT built upon this technique, utilizing gantry rotation in addition to the varying beam apertures for efficient treatment delivery.[4,5,6] For VMAT purposes, the CP concept became a “small arc approximation”,[1] with the apertures calculated at fixed gantry angle spacings and summed to produce the treatment arc.
The impact of CP spacing for VMAT treatment planning has been recognized as an important user-defined parameter previously for accurate and efficient VMAT treatment planning.[7,8] As it is a user-defined setting, a larger spacing between CP could potentially produce a suboptimal plan, along with greater differences between planned and measured doses, while reducing the spacing between CP may increase computational time.[7,9]
This study investigated the impact of CP discretization on dosimetry for a range of treatment plans, comparing two commercial TPSs. This treatment planning comparison was also validated by measurement-based verification.
METHODS
Treatment planning and delivery systems
A 6MV beam model of an Elekta linear accelerator (Elekta Oncology Systems, Crawley, UK) with an Agility MLC head modeled Pinnacle v16.21 (Philips Radiation Oncology Systems, Fitchburg, WI, USA) and RayStation v2023B (RaySearch Laboratories AB, Stockholm, Sweden) TPS, is considered for this study. The beam model in the Pinnacle TPS was commissioned for clinical VMAT planning at our center in February 2014 and was used for clinical treatment until August 2023. The beam model in the RayStation TPS was commissioned for clinical use in December 2020 and is currently being used for all clinical treatment planning. The dose calculation in the Pinnacle TPS is performed using the Adaptive Convolution algorithm, while the dose calculation in the RayStation TPS is performed using the Collapsed Cone Convolution algorithm. The beam models of both these systems are commissioned and validated as per TRS 430 guidelines.
The reference dosimetry of the linacs in our center is periodically audited by the Australian Clinical Dosimetry Service (ACDS, Melbourne, Australia) Level-I audit program and found to be at an optimal level. Similarly, the beam models of both Pinnacle and RayStation systems and the delivery of modulated treatment plans are audited by ACDS through Level-II and Level-III audits at the recommended frequency and are additionally found to be at an optimum level.[10,11,12]
Multi-leaf collimator model differences
The intra-leaf transmission and tongue and groove characteristics of MLC leaves are modeled and accounted for in dose calculation in a similar manner in both Pinnacle and RayStation. However, the rounded leaf tip design of MLC leaves is handled differently. In Pinnacle, the round leaf tip is managed by using off-axis distance-dependent position offset MLC leaves in the leaf movement direction.[13] In RayStation, the rounded leaf tip is primarily modeled by increasing transmission (the square of intra-leaf transmission) up to the user-defined leaf tip length and geometrical position offset in the leaf movement direction.[14,15]
Control points discretization for volumetric modulated arc therapy dose calculation
Pinnacle
In Pinnacle, VMAT dose calculation is divided into segments represented by CPs. Each CP is treated as a static beam, with three-dimensional (3D) dose calculation computed individually for CPs defining the dynamic VMAT plan delivery. CP weights are averaged for subsequent CPs to accommodate dynamic changes in beam parameters such as gantry angle and leaf position during delivery.
RayStation
In RayStation, VMAT dose calculation also involves segmenting the delivery into CPs. However, additional sub-segments are introduced between CPs based on the magnitude of leaf motion, ensuring leaf travel between sub-segments never exceeds 1.5 cm. Leaf positions and monitor units (MUs) are linearly interpolated between these sub-segments. Fluence from all segments, including sub-segments, is applied to the mean gantry position of subsequent CPs. Total energy released per unit mass (TERMA) is calculated for each segment by raytracing fluence into the patient and accumulated for each 10° section of the VMAT arc. The 3D dose deposited for each 10° arc section is computed by convolving the point dose kernel with the computed TERMA. Finally, the dose resulting from the VMAT arc is calculated by summing the doses from all arc sections.
Standard pattern volumetric modulated arc therapy arcs
Uniform fluence arcs
The study examined the impact of CP discretization on VMAT calculation by generating VMAT arcs using an in-house script that produces standard fluence patterns in the beam’s eye view. VMAT arcs with an arc length of 28° and a rectangular aperture sweeping from −12 cm to 12 cm were created in both Pinnacle and RayStation. Four aperture sizes (1 cm × 15 cm, 2 cm × 15 cm, 3 cm × 15 cm, and 4 cm × 15 cm) were considered. For each aperture size, arcs were generated with CP gantry spacings of 4° (resulting in 8 CPs) and 2° (resulting in 15 CPs), with equal weights assigned to all CPs. Figure 1a illustrates the resulting fluence in the beam’s eye view of the VMAT arc. Table 1 presents the leaf travel between successive CPs for the studied apertures for VMAT arcs with 2° and 4° CP spacing.
Figure 1.
(a) The uniform and (b) sine-modulated fluence generated by standard pattern volumetric modulated arc therapy arcs
Table 1.
Leaf travel distance between control points of volumetric modulated arc therapy arc with 2° and 4° control point gantry spacing, generating uniform fluence with different rectangular aperture sizes
| Aperture size | Leaf travel between subsequent CPs (cm) |
Leaf travel per degree (cm) | |
|---|---|---|---|
| 2° CP spacing | 4° CP spacing | ||
| 1 cm×15 cm | 1.53 | 3.29 | 0.765 |
| 2 cm×15 cm | 1.47 | 3.14 | 0.735 |
| 3 cm×15 cm | 1.40 | 3.0 | 0.7 |
| 4 cm×15 cm | 1.33 | 2.86 | 0.665 |
CP: Control points
Sine modulated volumetric modulated arc therapy fields
The relative weights of the CPs of uniform fluence VMAT arcs with 2° CP spacing and aperture sizes 1 cm × 15 cm, 2 cm × 15 cm, 3 cm × 15 cm, and 4 cm × 15 cm were adjusted using an in-house script to produce a sine-modulated fluence with four peaks and three troughs in the beam’s eye view. Figure 1b illustrates the electronic portal imaging device (EPID) measured fluence of the sine-modulated VMAT arc generated with a 1 cm × 15 cm aperture. Whilst the first three peaks of the generated fluence are identical in width and magnitude, the last peak has a larger width due to the residual weight of the arc after four peaks is assigned to the last CP. In this case, the weight assigned to the last CP is high, resulting in the broadening of the peak. The modulation was applied to arcs with CP spacings of 2° and aperture sizes of 1 cm × 15 cm, 2 cm × 15 cm, 3 cm × 15 cm, and 4 cm × 15 cm.
Measurement of standard pattern volumetric modulated arc therapy arcs
The uniform fluence and sine-modulated VMAT arcs were measured using the EPID available on the Elekta Versa HD Linac. The EPID images were converted to a planar dose image at a depth of 5.4 cm in a homogeneous water equivalent phantom with a 100 cm source-to-axis distance (SAD) geometry, utilizing a previously published EPID image dose conversion method.[16,17,18] The planar dose is calculated at a depth of 5.4 cm, as this methodology was validated against 2D ion chamber array-based measurements performed at the same depth in our institute. The calculated dose in EPID geometry from Pinnacle and RayStation using the collapsed VMAT arc QA methodology was compared using profile comparison with in-house developed software.[19,20] The measured and calculated dose profiles were quantitatively compared using the mean absolute percentage difference (MAPD) and the standard deviation of signed percentage differences (SDSPD). In the point-by-point comparison, signed percentage differences may include both positive and negative values, which can cancel when averaged and thereby underestimate the true magnitude of disagreement. To address this limitation, MAPD was used to quantify the overall magnitude of deviation independent of direction. In addition, SDSPD was used to characterize the spread of signed differences, providing a measure of variability around the mean signed agreement between measured and calculated profiles.
Patient dose calculation
Pinnacle plans calculated in RayStation
The VMAT plans generated in Pinnacle were recalculated using the beam model and dose engine available in RayStation to evaluate the consistency of the dose calculated by both systems. Figure 2a illustrates the study schema for comparing Pinnacle-generated plans calculated in RayStation. Ten prostate and 10 breast VMAT plans optimized with 4° CP spacing in Pinnacle were included. This CP spacing was chosen for clinical VMAT plan generation at our center to ensure practical optimization and plan generation times. The prostate and breast plans utilized dual full arc and dual partial arc VMAT techniques, respectively. All the plans were optimized with the leaf motion constraint value set to 4.6 mm/deg gantry angle.
Figure 2.
(a) The study schema of Pinnacle generated volumetric modulated arc therapy (VMAT) plans calculated in RayStation. (b) The study schema of RayStation generated VMAT plans calculated in Pinnacle. The solid lines indicate the schema of plans generated in the primary treatment planning system (TPS) and recalculated in an independent second TPS. The dashed line indicates the combination of dose calculations performed
To investigate the impact of CP spacing on patient dose calculation in Pinnacle, the dose calculations were performed for the optimized plans with 4° CP spacing, and CPs interpolated with 2° spacing using the “Interpolate” option available in Pinnacle. All plans generated in Pinnacle were imported into RayStation using the DICOM import functionality, and the dose was calculated on the patients’ CT images and RT structure datasets using RayStation’s beam model and dose engine. The consistency of the dose calculated by both systems was compared using dose volume histogram (DVH) metrics of the target volumes and organs at risk (OARs). Table 2 lists the target volumes and OARs along with their respective DVH metrics assessed for dose calculation comparison. The difference in DVH metrics calculated by the same planning system but with different CP spacings was assessed using the Wilcoxon signed-rank test. Similarly, the difference in DVH metrics between the calculations by Pinnacle and RayStation was also assessed using the Wilcoxon signed-rank test.
Table 2.
List of target volumes and organs at risk with corresponding dose-volume metrics assessed for dose calculated by Pinnacle and RayStation
| Treatment site | Structure | DVH metric |
|---|---|---|
| Prostate | CTV | Mean dose |
| Dose received by 99% volume (D99) | ||
| PTV | Mean dose | |
| Dose received by 95% volume (D95) | ||
| Dose received by 2% volume (D2) | ||
| Rectum | Mean dose | |
| Dose received by 60% volume (D60) | ||
| Dose received by 40% volume (D40) | ||
| Bladder | Mean dose | |
| Dose received by 50% volume (D50) | ||
| Dose received by 40% volume (D40) | ||
| Right femur | Mean dose | |
| Left femur | Mean dose | |
| Breast | CTV | Mean dose |
| Dose received by 99% volume (D99) | ||
| PTV | Mean dose | |
| Dose received by 95% volume (D95) | ||
| Dose received by 2% volume (D2) | ||
| Ipsi lateral lung | Mean dose | |
| Combined lung | Mean dose | |
| Heart | Mean dose | |
| Liver | Mean dose | |
| Dose received by 10 cc volume (D10cc) | ||
| Dose received by 60 cc volume (D60cc) | ||
| Esophagus | Dose received by 2 cc volume (D2cc) | |
| Spinal Canal | Dose received by 0.01 cc volume (D0.01cc) | |
| Trachea | Dose received by 0.01 cc volume (D0.01cc) |
CTV: Clinical target volume, PTV: Planning target volume, DVH: Dose volume histogram
RayStation plans calculated in Pinnacle
Similar to Pinnacle-generated plans, the VMAT plans created in RayStation were calculated using the beam model and dose engine available in the Pinnacle planning system. The CT and RT structure dataset were imported using DICOM import functionality, while the RT plans were imported using in-house software.[21] Figure 2b illustrates the study schema for comparing RayStation-generated plans calculated in Pinnacle. A cohort of 10 prostate and 10 breast patients was considered for this comparison. Clinically, a 2° CP spacing for VMAT plans is used with RayStation at our center, as plan optimization and dose calculation are faster, and this setting does not impact plan generation efficiency. All plans were optimized with the leaf motion constraint value set to 4.6 mm/deg gantry angle. Furthermore, in RayStation, there is no “Interpolate” option available to interpolate plans generated with coarse CP spacing to fine CP spacing. Hence, only the plans generated with 2° CP spacing in RayStation are compared with the dose calculation in Pinnacle. The RayStation and Pinnacle dose calculations are compared using the DVH metrics shown in Table 2. The Wilcoxon signed-rank test was used to assess the difference between RayStation and Pinnacle calculated DVH metrics. For the consistency in dose comparison purpose the RayStation calculated dose was considered as the reference dose in the determination for dose difference for both Pinnacle and RayStation generated plans.
Measured dose comparison
The prostate and breast VMAT plans generated in both Pinnacle and RayStation planning systems were evaluated using the ArcCheck helical diode array detector. The measured dose distributions were compared with the TPS-calculated dose on the digital image dataset of the ArcCheck phantom. For all Pinnacle-generated patient plans (dose calculated with 2° and 4° CP gantry spacing), ArcCheck QA plans were calculated in both Pinnacle and RayStation and compared against the measurements. Similar comparisons were performed for RayStation-generated plans.
The measured and TPS-calculated dose matrices were compared using 3% global dose tolerance and 3 mm distance-to-agreement (3%/3 mm) criteria, as well as 2%/2 mm tolerance criteria with a 10% low dose threshold.
RESULTS
Standard pattern volumetric modulated arc therapy arcs
Uniform fluence volumetric modulated arc therapy arc
Figure 3 illustrates planar dose calculations at a depth of 5.4 cm in a water-equivalent phantom by Pinnacle and RayStation for VMAT arcs with uniform fluence from a 1 cm × 15 cm aperture. Discretized dose calculations led to banding dose artifacts in both systems, with differences based on CP spacing and discretization methods. Pinnacle’s 4° CP spacing resulted in 8 distinct dose bands, halving the dose between adjacent bands [Figure 3a]. The 2° CP spacing showed double the bands [Figure 3b]. RayStation introduced sub-segments between CPs, with additional segments in the 4° arc to adhere to leaf travel constraints [Figure 3c]. The 2° arc also had sub-segments between CPs [Figure 3d].
Figure 3.
Pinnacle calculated planar dose at 5.4 cm depth of the homogeneous water equivalent phantom for a volumetric modulated arc therapy arc generating uniform fluence with the sweeping aperture of 1 cm × 15 cm and control point (CP) gantry spacing of (a) 4° and (b) 2°. RayStation calculated planar dose for the same field with CP gantry spacing of (c) 4° and (d) 2°
Figure 4 shows crossline profile comparisons for VMAT arcs with varying aperture sizes and CP spacings, alongside EPID-measured dose distributions. Table 3 summarizes the MAPD and the SDSPD between TPS-calculated and EPID-measured doses.
Figure 4.
The electronic portal imaging device measured and Pinnacle and RayStation calculated dose profiles for volumetric modulated arc therapy arcs calculated with 4° and 2° control point gantry spacing for the rectangular apertures of sizes (a) 1 cm × 15 cm, (b) 2 cm × 15 cm,(c) 3 cm × 15 cm, and (d) 4 cm × 15 cm
Table 3.
Mean absolute percentage difference and standard deviation of the signed percentage difference between the electronic portal imaging device, measured and treatment planning systems, calculated dose distributions for volumetric modulated arc therapy arcs generating uniform fluence across a range of aperture sizes, for both 4° and 2° control point gantry spacings
| Aperture size | MAPD (%) |
SDSPD (%) |
||||||
|---|---|---|---|---|---|---|---|---|
| Pinnacle | RayStation | Pinnacle | RayStation | |||||
|
| ||||||||
| 4° | 2° | 4° | 2° | 4° | 2° | 4° | 2° | |
| 1 cm×15 cm | 72.1 | 20.2 | 5.4 | 4.7 | 80.1 | 22.6 | 4.2 | 3.5 |
| 2 cm×15 cm | 38.3 | 9.7 | 2.4 | 9.8 | 42.0 | 10.5 | 3.6 | 11.1 |
| 3 cm×15 cm | 4.9 | 3.2 | 2.7 | 4.1 | 6.5 | 3.9 | 3.5 | 4.7 |
| 4 cm×15 cm | 16.7 | 2.4 | 4.3 | 2.9 | 17.9 | 3.3 | 4.9 | 3.9 |
MAPD: Mean absolute percentage difference, SDSPD: standard deviation of the signed percentage difference, CP: Control point
For the 1 cm × 15 cm aperture, Pinnacle 4° CP spacing exhibited large discrepancies (MAPD: 72.1%, SDSPD: 80.1%), which decreased substantially with 2° CP spacing (MAPD: 20.2%, SDSPD: 22.6%). In contrast, RayStation TPS showed much smaller differences for both 4° (MAPD: 5.4%, SDSPD: 4.2%) and 2° CP spacing (MAPD: 4.7%, SDSPD: 3.5%), indicating consistently better agreement with measurements.
For the 2 cm × 15 cm aperture, Pinnacle with 4° CP spacing showed higher discrepancies (MAPD: 38.3%, SDSPD: 42.0%), whereas RayStation demonstrated lower differences (MAPD: 2.4%, SDSPD: 3.6%). With 2° CP spacing, Pinnacle showed slightly improved agreement (MAPD: 9.7%, SDSPD: 10.5%) compared to RayStation (MAPD: 9.8%, SDSPD: 11.1%), with both systems performing comparably.
For the 3 cm × 15 cm aperture, discrepancies for Pinnacle at 4° CP spacing (MAPD: 4.9%, SDSPD: 6.5%) were markedly lower than those observed for smaller apertures. RayStation again showed lower differences (MAPD: 2.7%, SDSPD: 3.5%). With 2° CP spacing, Pinnacle demonstrated slightly better agreement (MAPD: 3.2%, SDSPD: 3.9%) compared to RayStation (MAPD: 4.1%, SDSPD: 4.7%).
For the 4 cm × 15 cm aperture, Pinnacle at 4° CP spacing showed higher discrepancies (MAPD: 16.7%, SDSPD: 17.9%), whereas RayStation maintained lower differences (MAPD: 4.3%, SDSPD: 4.9%). With 2° CP spacing, both systems showed good agreement, with Pinnacle (MAPD: 2.4%, SDSPD: 3.3%) performing slightly better than RayStation (MAPD: 2.9%, SDSPD: 3.9%).
Overall, RayStation demonstrated consistently better agreement with EPID measurements for 4° CP spacing across all aperture sizes. For 2° CP spacing, RayStation showed significantly better agreement for the smallest aperture (1 cm × 15 cm), while for larger apertures (≥2 cm × 15 cm), both systems showed comparable performance, with Pinnacle exhibiting marginally better agreement with measurements.
Sine-modulated volumetric modulated arc therapy arc
Figure 5a and b illustrate the planar dose calculated by Pinnacle and RayStation for a sine-modulated VMAT arc with a 1 cm × 15 cm aperture. Figure 5c and d depict similar data for a sine-modulated arc with a 2 cm × 15 cm aperture. The dose plane corresponds to the plane at a depth of 5.4 cm in a homogeneous phantom with a 100 cm SAD. Figure 6a-d presents the crossline dose profile comparison of the EPID measured dose, Pinnacle-calculated dose, and RayStation-calculated dose for VMAT arcs with the studied apertures.
Figure 5.
(a) Pinnacle and (b) RayStation calculated planar dose at 5.4 cm depth of the homogeneous water equivalent phantom for a volumetric modulated arc therapy (VMAT) arc generating sine modulated fluence with the sweeping aperture of 1 cm × 15 cm with 2° control point (CP) gantry spacing. (c) Pinnacle and (d) RayStation calculated planar dose at 5.4 cm depth of the homogeneous water equivalent phantom for a VMAT arc generating sine modulated fluence with the sweeping aperture of 2 cm × 15 cm with 2° CP gantry spacing
Figure 6.
The electronic portal imaging device measured, Pinnacle and RayStation calculated dose profiles for volumetric modulated arc therapy arcs generating Sine-modulated fluence with rectangular apertures (a) 1 cm × 15 cm, (b) 2 cm × 15 cm, (c) 3 cm × 15 cm, and (d) 4 cm × 15 cm
In all the studied apertures, Pinnacle-calculated dose exhibited dose artifacts, with the magnitude of the artifact reducing with increasing aperture size [Figures 5 and 6]. Conversely, the RayStation-calculated dose showed qualitative agreement with EPID measurements for both 1 cm × 15 cm and 4 cm × 15 cm arcs. However, for 2 cm × 15 cm and 3 cm × 15 cm arcs, RayStation exhibited artifacts similar to Pinnacle, but with mirrored peaks and valleys in the profile [Figures 5b, 6b, and c].
Table 4 presents the MAPD and SDSPD between the EPID measured and the TPS calculated dose profiles for the VMAT arc that produces sine modulated fluence. Overall, Pinnacle showed a larger difference with measurements compared to RayStation. For Pinnacle, the difference decreased with increasing aperture size. A similar trend was observed for RayStation; however, for the 2 cm × 15 cm aperture, the difference was relatively higher compared to the 1 cm × 15 cm aperture.
Table 4.
Mean absolute percentage difference and standard deviation of the signed percentage difference between electronic portal imaging device, measured and treatment planning systems, calculated dose distributions for volumetric modulated arc therapy arcs generating sine-modulated fluence across a range of aperture sizes
| Aperture size | MAPD (%) |
SDSPD (%) |
||
|---|---|---|---|---|
| Pinnacle | RayStation | Pinnacle | RayStation | |
| 1 cm×15 cm | 24.8 | 7.5 | 33.4 | 11.5 |
| 2 cm×15 cm | 13.6 | 10.6 | 19.7 | 12.6 |
| 3 cm×15 cm | 9.5 | 6.2 | 14.7 | 8.3 |
| 4 cm×15 cm | 7.3 | 5.1 | 11.9 | 6.9 |
MAPD: Mean absolute percentage difference, SDSPD: standard deviation of the signed percentage difference
Patient dose calculation comparison of Pinnacle and RayStation plans
Prostate plans
Table 5 displays the percentage difference in DVH metrics for target volumes and OARs between Pinnacle-generated prostate plans with 2° and 4° CP spacings, calculated in RayStation [as depicted in the study schema in Figure 2a]. Similarly, the corresponding data for RayStation-generated prostate plans with 2 CP spacings, calculated in Pinnacle, are presented in the same table, as illustrated in the study schema in Figure 2b.
Table 5.
The mean (standard deviation) difference in dose volume histogram metrics for target volumes and organs at risk between plans with different control point gantry spacing and Pinnacle-generated prostate volumetric modulated arc therapy plans calculated in RayStation, and vice versa
| Structure | DVH metric | Mean (SD) percentage difference between Pinnacle and RayStation calculated DVH metrics |
||||
|---|---|---|---|---|---|---|
| Pinnacle generated plans |
RayStation generated plans | |||||
| Pinnacle calculation | RayStation calculation | RayStation versus Pinnacle |
RayStation versus Pinnacle | |||
| 2° versus 4° | 2° versus 4° | 2° | 4° | 2° | ||
| CTV | Mean Dose | 0.2 (0.1)* | 0.0 (0.0) | 1.2 (0.4)* | 1.4 (0.4)* | 0.7 (0.3)* |
| D99 | 0.3 (0.4) | 0.1 (0.1)* | 0.4 (0.5)* | 0.6 (0.6)* | 1.0 (0.5)* | |
| PTV | Mean Dose | 0.4 (0.1)* | 0.0 (0.0) | 1.0 (0.4)* | 1.5 (0.4)* | 0.8 (0.2)* |
| D95 | 0.3 (0.2)* | 0.0 (0.0) | 0.7 (0.5)* | 0.9 (0.5)* | 1.0 (0.4)* | |
| D2 | 0.6 (0.4)* | 0.0 (0.0) | 1.4 (0.3)* | 2.0 (0.5)* | 4.2 (2.2)* | |
| Rectum | Mean Dose | −0.6 (0.2)* | 0.0 (0.0) | 1.4 (0.7)* | 0.8 (0.8)* | 0.0 (0.7) |
| D60 | −0.6 (0.8) | −0.1 (0.2) | 0.4 (1.2) | −0.1 (1.3) | −0.4 (1.4) | |
| D40 | −1.3 (0.4)* | −0.1 (0.2) | 2.3 (1.8)* | 1.1 (1.9) | −0.1 (1.7) | |
| Bladder | Mean Dose | −0.1 (0.3) | 0.0 (0.0) | −0.4 (1.1) | −0.5 (1.1) | 0.4 (0.7) |
| D50 | −0.1 (0.3) | 0.1 (0.2) | −2.1 (3.7) | −2.2 (3.8) | 0.4 (3.2) | |
| D40 | −0.2 (0.2) | −0.1 (0.1) | −0.1 (2.0) | −0.3 (1.9) | 3.3 (1.8)* | |
| Right Femur | Mean Dose | 0.0 (03) | −0.1 (0.1) | 0.9 (0.4)* | 1.0 (0.6)* | 0.5 (0.3)* |
| Left Femur | Mean Dose | 0.1 (0.2) | 0.0 (0.1) | 0.5 (0.5)* | 0.6 (0.5)* | 0.0 (1.7) |
*Statistically significant difference. CTV: Clinical target volume, PTV: Planning target volume, DVH: Dose volume histogram, SD: Standard deviation
There was a maximum planning target volume (PTV) D2 difference of 0.6 (0.4)% between 2° and 4° CP spacing plans calculated in Pinnacle. RayStation showed relatively less difference between 2° and 4° CP plans, with a maximum difference of 0.1 (0.1)% observed in clinical target volume D99. The 2° CP Pinnacle plans calculated in RayStation showed a maximum mean (Std Dev) difference of 1.4 (0.3)% for PTV D2. A similar comparison of 4° CP plans showed a maximum mean (Std Dev) difference of 2.0 (0.5)%. The RayStation plans calculated in Pinnacle showed a PTV mean dose difference of 0.8 (0.2)%. The D2 showed a mean (Std Dev) difference of 4.2 (2.2)%.
For the OARs dose, a maximum mean (Std Dev) difference of 3.3 (1.8)% was observed for Bladder D40 of RayStation plans calculated in Pinnacle.
Breast plans
The percentage difference of target volume and OARs DVH metrics between Pinnacle-generated and RayStation-calculated breast VMAT plans with 2° and 4° CP spacings is shown in Table 6. The corresponding data for RayStation-generated plans calculated in Pinnacle is also shown in the same table. Similar to the observations in prostate plans, the CP spacing has minimal impact on the target volumes and OARs dose in both Pinnacle and RayStation calculations.
Table 6.
Mean (standard deviation) of the difference in dose volume histogram metrics for target volumes and organs at risk between plans with different control point gantry spacing and Pinnacle-generated breast volumetric modulated arc therapy plans calculated in RayStation, and vice versa
| Structure | DVH metric | Mean (SD) percentage difference between Pinnacle and RayStation calculated DVH metrics |
||||
|---|---|---|---|---|---|---|
| Pinnacle generated plans |
RayStation generated plans | |||||
| Pinnacle calculation | RayStation calculation | RayStation versus Pinnacle |
RayStation versus Pinnacle | |||
| 2° versus 4° | 2° versus 4° | 2° | 4° | 2° | ||
| CTV | Mean dose | 0.4 (0.2)* | 0.0 (0.0)* | 0.2 (0.7) | 0.7 (0.8)* | 1.7 (0.4)* |
| D99 | −0.1 (0.2) | 0.0 (0.1) | 1.4 (1.5) | 1.3 (1.6)* | 9.2 (2.4)* | |
| PTV | Mean dose | 0.4 (0.2)* | 0.0 (0.0)* | 0.3 (0.8) | 0.8 (0.9)* | 2.4 (0.2)* |
| D95 | 0.0 (0.1) | 0.1 (0.1)* | 1.3 (1.0)* | 1.3 (1.0)* | 9.6 (1.1)* | |
| D2 | 0.9 (0.6)* | −0.1 (0.1)* | 0.6 (1.0) | 1.6 (1.6)* | 0.5 (0.4)* | |
| Ipsilateral Lung | Mean dose | 0.0 (0.1) | 0.0 (0.1)* | 0.4 (1.2) | 0.4 (1.2) | −0.1 (0.9) |
| Combined Lung | Mean dose | 0.0 (0.1) | 0.0 (0.1) | 0.0 (1.3) | 0.0 (1.3) | 2.9 (3.4)* |
| Heart | Mean dose | −0.4 (0.3)* | 0.0 (0.1) | −6.9 (2.2) | −7.3 (2.1)* | −3.9 (1.7)* |
| Liver | Mean dose | −0.3 (0.1)* | 0.0 (0.0) | −8.0 (12.4) | −8.3 (12.4) | −4.5 (2.3) |
| D10cc | 0.2 (0.6) | 0.0 (0.2) | 0.2 (1.4) | 0.3 (1.7) | −3.2 (2.5) | |
| D60cc | −0.7 (1.0) | −0.1 (0.1)* | −0.3 (3.0) | −0.9 (3.4) | −2.7 (1.9) | |
| Esophagus | D2cc | −0.4 (1.2) | 0.1 (0.3) | 2.5 (2.1)* | 1.9 (2.8) | 1.0 (1.1)* |
| Spinal Canal | D0.01cc | 1.2 (1.8) | −0.8 (1.0)* | 2.4 (2.2)* | 4.2 (3.4)* | −0.1 (1.2) |
| Trachea | D0.01cc | 0.5 (1.0) | −0.2 (0.4) | 2.9 (2.4)* | 3.6 (2.6)* | 2.8 (1.2)* |
*Statistically significant difference. CTV: Clinical target volume, PTV: Planning target volume, DVH: Dose volume histogram, SD: Standard deviation
The PTV D2 showed a maximum difference in target volume dose (mean (Std Dev) difference of 0.9 (0.6)%) between 2 and 4 CP spacing Pinnacle calculations. RayStation calculations showed relatively less difference, with a maximum mean (Std Dev) difference of 0.1 (0.1)% to PTV D95. Pinnacle-generated plans calculated in RayStation showed a mean (Std Dev) PTV mean dose difference of 0.3 (0.8)% and 0.8 (0.9)% for 2 and 4 CP spacing plans, respectively.
The RayStation plans calculated in Pinnacle showed the highest dose difference for the Target volumes and OARs DVH metrics. Figure 7 shows the relative dose difference on an axial slice for a representative Pinnacle-generated breast plan with 2° and 4° CP gantry spacing calculated in RayStation. The PTV mean dose and D2 showed a mean (Std Dev) difference of 2.4 (0.2)% and 9.6 (1.1)%, respectively. For the OAR dose, the 4 CP Pinnacle plans calculated in RayStation showed a maximum difference of −8.3 (12.4)%.
Figure 7.
Relative dose difference in an axial slice for a representative breast volumetric modulated arc therapy plan with 2° and 4° control point (CP) gantry spacing in Pinnacle and RayStation planning systems. (a) Dose difference between 2° and 4° CP spacing plans calculated in Pinnacle.(b) Dose difference between 2° and 4° CP spacing plans calculated in RayStation. (c) Dose difference between RayStation and Pinnacle 4° CP spacing plans. (d) Dose difference between RayStation and Pinnacle 2° CP spacing plans
ArcCheck measured dose comparison
The gamma analysis results of the ArcCheck measured dose with Pinnacle and RayStation calculated dose for prostate and breast VMAT plans are presented in Table 7a. Pinnacle-calculated breast and prostate plans with 4° CP spacing showed the lowest pass rates for the 2% 2 mm criteria compared to the respective 2° CP plans.
Table 7a.
Mean (standard deviation) gamma pass rate of ArcCheck measured and treatment planning systems-calculated dose of prostate and breast volumetric modulated arc therapy plans
| Control point gantry spacing | Mean (SD) Gamma pass rate (%) |
||||
|---|---|---|---|---|---|
| Prostate |
Breast |
||||
| 2%/2 mm | 3%/3 mm | 2%/2 mm | 3%/3 mm | ||
| Pinnacle generated plans | |||||
| Pinnacle calculated | 2° | 96.5 (1.2) | 99.7 (0.2) | 91.5 (5.2) | 97.9 (2.5) |
| 4° | 93.1 (2.2) | 98.8 (0.5) | 88.9 (4.0) | 96.3 (3.0) | |
| RayStation calculated | 2° | 96.3 (0.9) | 99.7 (0.3) | 90.3 (5.6) | 98.2 (1.6) |
| 4° | 95.6 (1.1) | 99.5 (0.3) | 89.9 (4.8) | 97.5 (2.2) | |
| RayStation generated plans | |||||
| Pinnacle calculated | 2° | 96.8 (1.4) | 99.5 (0.4) | 95.4 (2.1) | 99.5 (0.3) |
| RayStation calculated | 2° | 96.7 (2.3) | 99.6 (0.5) | 95.4 (3.7) | 99.6 (0.6) |
SD: Standard deviation
The mean (Std Dev) difference in gamma pass rate of measured and TPS calculated dose analysis between 2° and 4° CP spacing plans calculated in RayStation and Pinnacle is shown in Table 7b. The same data for 2° and 4° CP plans between RayStation and Pinnacle, calculated dose for Pinnacle-generated plans are also shown in the same table. Similar data for RayStation-generated plans calculated in Pinnacle are also presented in Table 7b.
Table 7b.
Mean (standard deviation) of gamma pass rate difference between ArcCheck measured and treatment planning systems calculated dose for plans with different control point gantry spacing and Pinnacle-generated volumetric modulated arc therapy plans calculated in RayStation and vice versa
| Treatment site | Gamma tolerance criteria | Mean (SD) Gamma pass rate difference (%) |
||||
|---|---|---|---|---|---|---|
| Pinnacle generated plans |
RayStation generated plans | |||||
| Pinnacle calculation | RayStation calculation | RayStation versus Pinnacle |
RayStation versus Pinnacle | |||
| 2° versus 4° | 2° versus 4° | 2° | 4° | 2° | ||
| Prostate | 2%/2 mm | 3.4 (1.3)* | 0.7 (0.7)* | −0.2 (0.9) | 2.5 (1.9)* | −0.1 (1.2) |
| 3%/3 mm | 0.8 (0.4)* | 0.2 (0.2)* | 0.0 (0.4) | 0.7 (0.7)* | 0.1 (0.4) | |
| Breast | 2%/2 mm | 3.6 (1.5)* | 1.0 (0.7)* | −0.9 (2.7) | 1.2 (1.5) | 0.0 (2.4) |
| 3%/3 mm | 16 (1.2)* | 0.4 (0.6)* | 0.1 (0.9) | 1.7 (3.2)* | 0.1 (0.5) | |
*Statistically significant difference. SD: Standard deviation
There was a statistically significant difference in gamma pass rate between 2° and 4° CP Pinnacle-generated plans in both RayStation and Pinnacle calculations. In addition, 4° CP plans generated in Pinnacle showed a statistically significant difference in pass rate between the measurement and the RayStation and Pinnacle-calculated dose.
DISCUSSION
In current clinical radiotherapy planning systems, dose calculation for dynamic radiotherapy delivery is approximated using multiple equidistant static beams. While increasing the number of static beams can intuitively improve dose calculation accuracy, a trade-off must be made to ensure optimization and dose calculation times remain practical without compromising accuracy. Various strategies have been implemented to optimize the optimization time of VMAT plans. For instance, in RapidArc plan generation using Eclipse, optimization begins with 5 equidistant CPs for a full arc VMAT plan and progressively adds more CPs during optimization, with a final gantry spacing of 2°.[22] In Pinnacle and RayStation, optimization starts with fluence-based optimization of CPs defined at 24° angular separation over the defined arc in the first level of optimization, followed by introducing user-defined gantry spacing CPs between the initial ones for further optimization.[23] While these approaches affect optimization time and plan quality to some extent, the accuracy of the generated plan primarily depends on the final resolution of the CPs and leaf travel distance between subsequent CPs.[8,24,25]
In one of the earliest VMAT optimization papers, Otto postulated that a CP gantry angle spacing of 1° and leaf motion not exceeding 5 mm per gantry angle are necessary to achieve dose calculation accuracy in VMAT planning.[2] However, the majority of commercial planning systems offer user-defined CP spacing in the range of 2°–6°, along with user-defined leaf motion constraint parameters. Both CP gantry spacing and leaf travel per gantry angle play vital roles in determining dose calculation accuracy in VMAT.
Feygelman et al. studied the dosimetric accuracy of VMAT plans generated for TG119 test cases with 2°, 4°, and 6° CP spacing in Pinnacle and concluded that plans with 4° CP spacing provided acceptable dosimetric accuracy with reasonable optimization and dose calculation time.[7] Generally, studies have shown acceptable dose accuracy with 2° CP spacing in RapidArc and 4° CP spacing in Pinnacle plans.
In this study, we examined the VMAT CP discretization implemented in both Pinnacle and RayStation using VMAT arcs that generate standard fluence, as well as clinical prostate and breast VMAT plans. The primary aim of utilizing standard fluence VMAT arcs was to assess the impact of CP discretization on dose calculation in both systems, with a focus on CP spacing, which determines the leaf travel distance between CPs (also known as leaf travel per degree of gantry rotation).
The introduction of sub-segments between CPs by RayStation significantly improved dose agreement between measurement and calculation for both uniform and sine-modulated VMAT arcs, compared to Pinnacle-calculated doses, for both 2° and 4° CP spacing. This improvement is expected to enhance the accuracy of patient dose calculations. In Pinnacle calculations, the 2° CP arc consistently exhibited better agreement with measurements compared to 4° CP plans, likely due to the increased number of static arcs representing dynamic delivery.
In RayStation calculations, the agreement between measured and calculated doses was primarily determined by the leaf travel distance between CPs rather than CP spacing. RayStation defines a leaf travel threshold of a minimum of 1.5 cm between CPs for the introduction of interpolated sub-segments. This approach may yield better dose agreement for VMAT arcs with coarse CP spacing, where the leaf travel between subsequent CPs exceeds the threshold, resulting in the introduction of multiple sub-segments. However, for arcs with fine CP spacing, dose discrepancies may arise due to the absence of sub-segments, as the leaf travel is below the threshold.
For example, in a uniform fluence VMAT arc with a 2 cm × 15 cm aperture, the 4° VMAT arc in RayStation demonstrated improved agreement with measured dose compared to its equivalent 2° arc. This was primarily due to the 4° arc resulting in a leaf travel of 3.14 cm between successive CPs, leading to the addition of three additional sub-segments between CPs. Consequently, 29 static beams (comprising 21 sub-segments and 8 CPs) were used for the arc to maintain leaf travel of <1.5 cm between adjacent segments. In contrast, the 2° arc had a leaf travel between successive CPs of 1.47 cm, which was less than the leaf travel threshold required for the introduction of sub-segments. Therefore, the arc was effectively calculated with 15 static beams (defined by CPs). Implementing sub-segments based on leaf travel distance per degree rather than per CP in RayStation would likely result in more consistent dose calculations. The dose profile for uniform fluence calculated using a 3 cm × 15 cm aperture with 4° CP spacing shows better agreement with the measurements compared to 2° CP spacing for both Pinnacle and RayStation calculations [Figure 4c]. This is attributed to the favorable alignment of the aperture edges in the leaf motion direction defined by the 8 CPs. For the same reason, the dose calculated for the 4 cm × 15 cm aperture with 2° CP spacing by both planning systems also shows close agreement with the EPID measurements [Figure 4d].
In general, at coarser angular resolution (4° CP spacing), differences in how the two systems model leaf motion and dose interpolation between CPs can lead to larger discrepancies, particularly for larger leaf motion between CPs. With finer sampling (2° CP spacing), the discretization error is reduced, allowing both systems to better approximate continuous delivery, thereby minimizing differences. At the finer CP spacing, Pinnacle calculation agreed marginally better for the uniform fluence VMAT arc for the aperture size ≥2 cm × 15 cm than RayStation [Table 3]. However, in sine modulated VMAT arcs, RayStation consistently showed better agreement with the measurement than Pinnacle [Table 4]. For the 2 cm × 15 cm sine-modulated arc, the differences observed with RayStation are higher than those for the 1 cm × 15 cm aperture. This is likely due to the introduction of sub-segments in the 1 cm × 15 cm case, where leaf travel between CPs exceeds 1.5 cm. These additional sub-segments improve the approximation of continuous delivery and reduce discretization error. In contrast, for the 2 cm × 15 cm aperture, only the CP-defined segments are used, resulting in comparatively higher discretization effects and dose differences.
In clinical VMAT plans, Pinnacle dose calculations exhibited relatively larger differences between 2° and 4° CP plans. For breast plans, the mean dose and D2 to PTV showed a mean (SD) difference of 0.4 (0.2) % and 0.9 (0.6) %, respectively, and these differences were statistically significant. The 2° CP plan resulted in higher doses compared to the 4° CP plan.
Chen et al. investigated dose variations in target volumes and OARs in prostate VMAT plans optimized with 10° CP and then recalculated with linearly interpolated 2° CP spacing. They also explored the effect of the leaf motion constraint in both scenarios.[9] Their results indicated that when the leaf motion was constrained to 2 mm/deg, the PTV and rectum doses agreed within 1.4% for both 10° and 2° CP plans. However, with a relaxed leaf motion constraint of 20 mm/deg, dramatic differences in dose to target volumes and OARs were observed between the 10° and 2° CP plans. Specifically, the mean dose to PTV increased by 5.3% in the 2° CP plan compared to the 10° CP plan, with a 150% increase in the standard deviation of PTV dose, indicating dose heterogeneity within the PTV. Similarly, the mean dose to the rectum and bladder increased by 20% and 7.7%, respectively, in the 2° CP plans compared to the 10° CP plan.
Our study revealed a similar trend in mean dose to target volumes between 2° and 4° CP prostate and breast plans calculated in Pinnacle. However, the magnitude of dose differences observed was small, as the plans were generated with a leaf motion constraint of 0.46 cm/deg and relatively less difference between the CP gantry spacing among the studied cohorts.
The DVH metrics difference between 2° and 4° CP plans calculated in RayStation showed good agreement, with a maximum mean (Std Dev) difference of − 0.1 (0.1) % observed for PTV D2 in breast plans. Understandably, the fixed number of static beams representing CPs resulted in dose differences between 2° and 4° CP plans in Pinnacle. In contrast, the sub-segments approach implemented in RayStation resulted in additional static beams and less dose difference between 2° and 4° CP plans.
The 4° CP Pinnacle-generated prostate plans recalculated in RayStation exhibited a maximum target dose difference between the calculations of the same plan in two systems. RayStation calculation resulted in a higher PTV mean dose, with a mean (Std Dev) difference of 1.5 (0.4)%, compared to Pinnacle calculations. The 2° CP plans showed a similar trend with a relatively reduced dose difference. The reduction in dose difference with reduced CP spacing is attributed to the better approximation of the arc by static beams with high-resolution CPs in Pinnacle. Overall, Pinnacle-generated plans calculated in RayStation showed a consistently higher mean dose, by 0.7 (0.6)%, to the target volumes for combined prostate and breast plans with 2° CP spacing.
The difference in OARs DVH metrics between Pinnacle and RayStation also reduced in 2° CP plans compared to 4° CP plans for both studied treatment sites. The agreement of PTV D2 and D0.01cc of the trachea and spinal canal, parameters that represent the maximum dose in the volumes, between Pinnacle and RayStation calculations improved in 2° CP plans. However, the mean dose to the heart and liver showed relatively higher percentage differences between the doses calculated by both systems. These structures are at the edge of the treatment field and receive a mean dose in the range of 3% to 12% of the target prescription dose. The observed dose differences are in the range of 0.3% to 1.5% of the prescription dose. Nevertheless, the dose calculated by RayStation for these structures was systematically lower compared to Pinnacle calculations. A small difference in out-of-field beam modeling between Pinnacle and RayStation could be the reason for this observed systematic difference.
The RayStation-generated prostate plans calculated in Pinnacle showed better agreement with the mean dose to target volumes compared to the Pinnacle-generated prostate plans calculated in RayStation [Table 5]. In contrast, RayStation-generated breast plans calculated in Pinnacle showed higher dose differences compared to the 2° CP Pinnacle breast plans calculated in RayStation. A similar trend was observed for OAR dose agreement for both prostate and breast plans.
Overall, the target volume dose calculated by RayStation was generally higher than that calculated by Pinnacle, irrespective of whether the plans were generated in Pinnacle or RayStation. The largest dose differences were observed for breast plans optimized in RayStation. The observed differences between Pinnacle and RayStation calculations may be partly attributed to variations in VMAT segment complexity; however, this was not explicitly quantified in the present study and warrants further investigation. In particular, the relatively complex segments associated with breast VMAT plans optimized in RayStation may contribute to the observed discrepancies when recalculated in Pinnacle, as the MLC modeling parameters in Pinnacle are primarily tuned for plans generated within its own optimization framework. In contrast, Pinnacle-optimized plans recalculated in RayStation showed comparatively better agreement, suggesting that the RayStation dose calculation model is more robust to variations in segment characteristics originating from Pinnacle-generated plans.
The ArcCheck gamma analysis of measured dose against TPS-calculated dose shows that with a 2%/2 mm tolerance criterion, the plans calculated with 4° CP spacing in RayStation showed better agreement with measurements compared to Pinnacle calculations, supporting the increased dosimetric accuracy with the introduction of sub-segments in the 4° CP plans calculated in RayStation. Increased dosimetric accuracy with high-resolution CPs has been demonstrated previously by other studies.[2,7,26,27] While increasing the sub-segments better approximates the dynamic delivery, it does not model all the characteristics of dynamic treatments, such as acceleration and deceleration of leaf and gantry motion, and ramp-up and ramp-down of dose rate during the change in dose rate. Boylan et al. modeled the leaf and gantry motion characteristics, including the gravity effect on these parameters, of the VMAT delivery in Monte Carlo-based dose calculation and found improved agreement of the calculated dose with the measured dose.[28] In commercial clinical planning systems, the leaf latency of binary leaves is accounted for in the Tomotherapy dose calculation;[29] however, aspects of dynamic VMAT delivery are yet to be implemented for improved dose calculation accuracy.
In this work, the clinical beam models used in two commercial planning systems that have two distinct approaches to model VMAT delivery have been studied. Our results suggest that in clinical prostate and breast VMAT plans, the plans generated with 2° CP spacing showed minimal dose difference between them, which is comparable to the results of other studies.
CONCLUSIONS
The CP discretization approaches implemented in the Pinnacle and RayStation planning systems are characterized, and the dosimetric differences arising from these two different approaches are studied. Pinnacle exhibited dramatic dose differences for the collapsed arc dose calculations of the VMAT arcs, generating uniform and sine-modulated fluence, while RayStation calculated doses for these arcs closely matched the measured dose. Pinnacle-generated VMAT plans calculated in RayStation demonstrated that the VMAT dose calculation accuracy in RayStation is less affected by CP spacing. However, in Pinnacle, improved dose calculation accuracy is achieved with 2° CP gantry spacing compared to 4° CP gantry spacing. Small dose differences to the target volume were observed between Pinnacle and RayStation for Pinnacle-generated prostate and breast VMAT plans; however, the breast VMAT plans generated in RayStation showed a higher PTV mean dose difference, up to 2.4 (0.2)%, compared to the same plans calculated in Pinnacle. The sub-segment-based VMAT dose calculation approach implemented in RayStation results in improved dose calculation accuracy of VMAT plans compared to the Pinnacle dose calculation.
Conflicts of interest
There are no conflicts of interest.
Funding Statement
Nil.
REFERENCES
- 1.Webb S, McQuaid D. Some considerations concerning volume-modulated arc therapy: A stepping stone towards a general theory. Phys Med Biol. 2009;54:4345–60. doi: 10.1088/0031-9155/54/14/001. [DOI] [PubMed] [Google Scholar]
- 2.Otto K. Volumetric modulated arc therapy: IMRT in a single gantry arc. Med Phys. 2008;35:310–7. doi: 10.1118/1.2818738. [DOI] [PubMed] [Google Scholar]
- 3.Webb S. Optimisation of conformal radiotherapy dose distributions by simulated annealing. Phys Med Biol. 1989;34:1349–70. doi: 10.1088/0031-9155/34/10/002. [DOI] [PubMed] [Google Scholar]
- 4.Yu CX. Intensity-modulated arc therapy with dynamic multileaf collimation: An alternative to tomotherapy. Phys Med Biol. 1995;40:1435–49. doi: 10.1088/0031-9155/40/9/004. [DOI] [PubMed] [Google Scholar]
- 5.Bortfeld T. Optimized planning using physical objectives and constraints. Semin Radiat Oncol. 1999;9:20–34. doi: 10.1016/s1053-4296(99)80052-6. [DOI] [PubMed] [Google Scholar]
- 6.Webb S. The physical basis of IMRT and inverse planning. Br J Radiol. 2003;76:678–89. doi: 10.1259/bjr/65676879. [DOI] [PubMed] [Google Scholar]
- 7.Feygelman V, Zhang G, Stevens C. Initial dosimetric evaluation of SmartArc – A novel VMAT treatment planning module implemented in a multi-vendor delivery chain. J Appl Clin Med Phys. 2010;11:3169. doi: 10.1120/jacmp.v11i1.3169. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Bzdusek K, Friberger H, Eriksson K, Hårdemark B, Robinson D, Kaus M. Development and evaluation of an efficient approach to volumetric arc therapy planning. Med Phys. 2009;36:2328–39. doi: 10.1118/1.3132234. [DOI] [PubMed] [Google Scholar]
- 9.Chen F, Rao M, Ye JS, Shepard DM, Cao D. Impact of leaf motion constraints on IMAT plan quality, deliver accuracy, and efficiency. Med Phys. 2011;38:6106–18. doi: 10.1118/1.3651698. [DOI] [PubMed] [Google Scholar]
- 10.Williams I, Kenny J, Lye J, Lehmann J. Berlin/Heidelberg, Germany: Springer; 2012. World Congress on Medical Physics and Biomedical Engineering May 26-31, 2012, Beijing, China; pp. 1215–8. [Google Scholar]
- 11.Lehmann J, Alves A, Dunn L, Shaw M, Kenny J, Keehan S, et al. Dosimetric end-to-end tests in a national audit of 3D conformal radiotherapy. Phys Imaging Radiat Oncol. 2018;6:5–11. doi: 10.1016/j.phro.2018.03.006. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Available from: https://www.arpansa.gov.au/our-services/testing-and-calibration/australian-clinical-dosimetry-service/acds-audit-program . [Last accessed on 2025 Jun 13]
- 13.Young LA, Yang F, Cao N, Meyer J. Rounded leaf end modeling in pinnacle VMAT treatment planning for fixed jaw linacs. J Appl Clin Med Phys. 2016;17:149–62. doi: 10.1120/jacmp.v17i6.6343. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Saez J, Hernandez V, Goossens J, De Kerf G, Verellen D. A novel procedure for determining the optimal MLC configuration parameters in treatment planning systems based on measurements with a Farmer chamber. Phys Med Biol. 2020;65:155006. doi: 10.1088/1361-6560/ab8cd5. [DOI] [PubMed] [Google Scholar]
- 15.Hernandez V, Angerud A, Bogaert E, Hussein M, Lemire M, García-Miguel J, et al. Challenges in modeling the agility multileaf collimator in treatment planning systems and current needs for improvement. Med Phys. 2022;49:7404–16. doi: 10.1002/mp.16016. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Alharthi T, Arumugam S, Vial P, Holloway L, Thwaites D. EPID sensitivity to delivery errors for pre-treatment verification of lung SBRT VMAT plans. Phys Med. 2019;59:37–46. doi: 10.1016/j.ejmp.2019.02.007. [DOI] [PubMed] [Google Scholar]
- 17.Lee C, Menk F, Cadman P, Greer PB. A simple approach to using an amorphous silicon EPID to verify IMRT planar dose maps. Med Phys. 2009;36:984–92. doi: 10.1118/1.3075817. [DOI] [PubMed] [Google Scholar]
- 18.Xing A, Arumugam S, Deshpande S, George A, Holloway L, Goozee G, et al. Streamlining EPID-based IMRT quality assurance: auto-analysis and auto-report generation. Journal of Physics: Conference Series. 2014;489:012084. [Google Scholar]
- 19.Arumugam S, Xing A, Jameson M, Holloway L. An algorithm to calculate a collapsed arc dose matrix in volumetric modulated arc therapy. Med Phys. 2013;40:071724. doi: 10.1118/1.4810964. [DOI] [PubMed] [Google Scholar]
- 20.Young T, Xing A, Vial P, Thwaites D, Holloway L, Arumugam S. Sensitivity of collapsed arc QA method for delivery errors in Volumetric Modulated Arc Therapy (VMAT) Journal of Physics: Conference Series. 2015;573:012021. [Google Scholar]
- 21.Arumugam S, Xing A, Goozee G, Holloway L. Independent calculation-based verification of IMRT plans using a 3D dose-calculation engine. Med Dosim. 2013;38:376–84. doi: 10.1016/j.meddos.2013.04.005. [DOI] [PubMed] [Google Scholar]
- 22.Varian Medical Systems, Inc; Palo Alto, CA, United States of America: 2015. Eclipse Photon and Electron Algorithms Reference Guide; pp. 169–71. [Google Scholar]
- 23.RaySearch Laboratories; Stockholm, Sweden: VMAT optimization in RayStation- White Paper. [Google Scholar]
- 24.Papp D, Unkelbach J. Direct leaf trajectory optimization for volumetric modulated arc therapy planning with sliding window delivery. Med Phys. 2014;41:011701. doi: 10.1118/1.4835435. doi: 10.1118/1.4835435. [DOI] [PubMed] [Google Scholar]
- 25.Pardo Montero J, Fenwick JD. The effect of different control point sampling sequences on convergence of VMAT inverse planning. Phys Med Biol. 2011;56:2569–83. doi: 10.1088/0031-9155/56/8/015. [DOI] [PubMed] [Google Scholar]
- 26.Arumugam S, Berry M, Ochoa C, Xing A, Beeksma B, Vial P. Dose discrepancy between planning system estimation and measurement in spine stereotactic body radiation therapy: A case report. J Med Imaging Radiat Oncol. 2017;61:695–8. doi: 10.1111/1754-9485.12617. [DOI] [PubMed] [Google Scholar]
- 27.Masi L, Doro R, Favuzza V, Cipressi S, Livi L. Impact of plan parameters on the dosimetric accuracy of volumetric modulated arc therapy. Med Phys. 2013;40:071718. doi: 10.1118/1.4810969. doi: 10.1118/1.4810969. [DOI] [PubMed] [Google Scholar]
- 28.Boylan CJ, Aitkenhead AH, Rowbottom CG, Mackay RI. Simulation of realistic linac motion improves the accuracy of a Monte Carlo based VMAT plan QA system. Radiother Oncol. 2013;109:377–83. doi: 10.1016/j.radonc.2013.08.046. [DOI] [PubMed] [Google Scholar]
- 29.Westerly DC, Soisson E, Chen Q, Woch K, Schubert L, Olivera G, et al. Treatment planning to improve delivery accuracy and patient throughput in helical tomotherapy. Int J Radiat Oncol Biol Phys. 2009;74:1290–7. doi: 10.1016/j.ijrobp.2009.02.004. [DOI] [PMC free article] [PubMed] [Google Scholar]