Technical Perspectives | Eclipse Photon Beam Models for Halcyon
ECLIPSE PHOTON BEAM MODELS FOR HALCYON
Stephen Thompson1, Magdalena Constantin1, Petri Kokkonen2
Introduction
A ring design system (RDS) treatment delivery device, the
Halcyon™ system is supported in the Eclipse™ treatment
planning system with photon beam models for the
anisotropic analytical algorithm (AAA) and the photon
optimization (PO) algorithm. Embedded within AAA is
support for both portal dose and MV imaging dose. The
imaging dose is calculated during both optimization and
final dose calculations with a newly developed Fourier
transform dose calculation (FTDC) algorithm. Details of the
AAA, AAA for portal dose, and PO algorithms are available
in the Eclipse Photon and Electron Algorithms Reference
Guide. The FTDC component of AAA is described in the
RDS Treatment Planning Instructions for Use document.
Beam configuration
The AAA and PO photon beam models for Halcyon are
pre-configured and are not customizable by the user. The
beam configuration for each algorithm has been performed
by the Varian Applied Research team using a combination
of automated and manual optimizations, minimizing the
gamma error between the measured and calculated profiles
and percent depth doses (PDDs). The beam data set used
for algorithm configuration was acquired at the factory3
with the hardware and parameters described in Appendix A.
Beam model parameters were optimized such that the
model calculated profiles and PDDs are within our internal
specified accuracies (i.e., for open 6X flattening filter free
(FFF) beams, > 90% of points calculated within water
equivalent media are within 2%, 2 mm gamma in both high
and low dose regions, and maximum gamma < 2.5 mm for
all measured points). Accuracy requirements for the three
dose calculation algorithms are shown in Table 1.4
As shown in Table 1, the internally validated calculation
accuracy requirements for the FTDC algorithm, which
calculates MV-MV and MV-CBCT doses, are within a
gamma of 5%/5 mm. The relatively high gamma values are
intentional so that the coarser calculation grid and known
uncertainties (i.e., positioning) do not negatively influence
the overall planning and delivery qualities.
This document contains confidential, proprietary information of Varian Medical Systems. It may not be copied or reproduced without prior
written permission from Varian Medical Systems.
Gamma Pass Maximum
Algorithm
Criteria Rate Gamma
AAA* 2%, 2 mm >90% 100% <2.5
FTDC* 5%, 5 mm >90% 100% <2.5
AAA Portal Dose** 3%, 3 mm >95% 99% < 1.5
Table 1. Internal accuracy requirements for dose calculation
algorithms with the RDS photon beam model.
* Specified accuracy is in water equivalent media.
** Specified accuracy of predicted image is within the electronic portal
imaging device (EPID) media.
For AAA, the gamma error histogram in the Eclipse
Beam Configuration can be inspected to evaluate the
correspondence of the measured and calculated profiles
and PDDs of the beam model. The Eclipse Photon and
Electron Algorithms Reference Guide and Eclipse Beam
Configuration Reference Guide provide additional details
for gamma error tolerances for AAA. For the multi-
resolution dose calculation (MRDC) within the photon
optimization algorithm, the maximum error for dose
estimation was calculated during configuration and is
shown in the Beam Configuration workspace.
Figures 1 through 6 depict the correspondence of the
calculated and measured profiles and PDDs for the RDS
device in Eclipse software version 15.1.1 (Distributed
calculation framework (DCF) build 15.1.51) for the 2 x 2 cm2,
10 x 10 cm2, and 28 x 28 cm2 fields. The 3D dose matrices
(from which the profiles and PDD are extracted) were
normalized such that 100% is on the central axis (CAX) at 5
cm deep. These figures were calculated from Varian internal
automated regression tests and will have differences from
what is observed in the Eclipse Beam Analysis module.
However, users can utilize the Beam Analysis to calculate
PDD and profiles for all measured field sizes and depths. The
Beam Analysis tools can then be used to further assess the
agreement between measured and calculated beam profiles
and PDDs in Eclipse. See Analyzing and Approving Beam
Data in Chapter 5 of Eclipse Beam Configuration Reference
Guide for additional information.
 Figure 1. AAA 15.1.51 2 x 2 cm2 depth doses (measured versus Figure 4. AAA 15.1.51 2 x 2 cm2 crossline profiles (leaf motion direction)
calculated) from the Varian internal comparison framework. at 10 cm depth from the Varian internal comparison framework.

Figure 2. AAA 15.1.51 10 x 10 cm2 depth doses (measured versus Figure 5. AAA 15.1.51 10 x 10 cm2 crossline profiles (leaf motion direction)
calculated) from the Varian internal comparison framework. at 10 cm depth from the Varian internal comparison framework.

Figure 3. AAA 15.1.51 28 x 28 cm2 depth doses (measured versus Figure 6. AAA 15.1.51 28 x 28 cm2 crossline profile (leaf motion direction)
calculated) from the Varian internal comparison framework. at 10 cm depth from the Varian internal comparison framework. Notice
that the water phantom used in calculation is 40 cm wide and therefore
does not extend beyond the [-200,200] mm interval.

This document contains confidential, proprietary information of Varian Medical Systems. It may not be copied or reproduced without prior
written permission from Varian Medical Systems.
2
 Technical Perspectives | Eclipse Photon Beam Models for Halcyon

Representative screenshots from beam analysis for the Dosimetric parameters

10 x 10 cm2 field size are shown in Figures 7 and 8.
The AAA beam model uses the externally visible dosimetric
parameters shown in Table 2. The per layer leaf transmission
was measured with a Farmer chamber at isocentric depths
from 5 cm to 20 cm for both proximal and distal leaf banks.
Results for each bank are nearly identical (within 0.2%)
and a single value is used in the beam model. Due to beam
hardening, a range of 2.3% in the leaf transmission was
observed over isocentric depths from 5 cm to 20 cm. We
have selected a value which provides good results in a wide
variety of clinical cases.

Leaf Transmission Dosimetric AAA Effective

(per layer) Leaf Gap Spot Size

0.47% 0.1 mm (0, 0)

Table 2. The leaf transmission, dosimetric leaf gap, and AAA

Figure 7. AAA 15.1.51 10 x 10 cm2 depth doses (measured versus
effective spot size for the AAA 15.1.51 beam model
calculated) from the Eclipse Beam Analysis module.
Representative measurements of leaf transmission per
layer and per bank for various source-to-surface distance
(SSD) and depths are shown in Table 3.

The MLC fluence is modeled in Eclipse with additional

parameters including tongue-and-groove (TnG) and interleaf
leakage. Values used for these parameters have been tuned
to provide good agreement for clinical scenarios. The TnG
used internally is on the order of 0.3 mm. Small changes to
tune the observed profiles to measurements have been used
but are within 0.05 mm of each other.

The Interleaf leakage was measured both using a water tank

(90 cm SSD, 10 cm depth) and solid water with Gafchromic™
EBT3 film. The leaves from a single bank in a single layer
configuration were fully extended across the field and
Figure 8. AAA 15.1.51 10 x 10 cm2 crossline profiles at 10 cm depth
the secondary layer of leaves was retracted. For the water
(measured versus calculated) from the Eclipse Beam Analysis module.
scans, the Sun Nuclear Corporation EDGE Detector™ diode
was used. The maximum interleaf leakage for the stepped
MLC design was found to be 0.7% relative to the reference
10 cm open field.

Average Average
SSD [cm] Depth [cm] Prox. A Prox. B Dist. A Dist. B (A,B)-Prox. (A,B)-Dist.

95 5 .45% .44% .46% .45% .45% .45%

90 10 .49% .48% .50% .49% .49% .49%

Table 3. Leaf transmission measurements (per layer, per bank) using the PTW Farmer ionization chamber. The proximal leaves (“Prox.”) are
closer to the source, and the distal leaves (“Dist.”) are further away from the source. A and B refer to bank A and bank B leaves.

This document contains confidential, proprietary information of Varian Medical Systems. It may not be copied or reproduced without prior
written permission from Varian Medical Systems.
3
 Figure 9 shows representative measurement data for
interleaf leakage. Calculations from the Eclipse AAA are
shown compared to measurements in Figure 10.
The dosimetric leaf gap (DLG) value in Table 2 corresponds
to a single layer scenario. Experimental measurements using
the sweeping gap test indicated that proximal and distal
leaves have very similar DLGs in the 100 to 200 micron
range. Figure 11 presents the DLG measurements for the
distal leaves.

Figure 9. Screen shot from OmniPro-Accept 7.5 of water scans for interleaf leakage test for the distal leaves. Different scans correspond to
different plans: distal A interleaf, distal B interleaf and full interdigitation interleaf leakage. For reference, the maximum open field profile is
shown, scaled to the average leaf transmission level (~0.5%).

Figure 11. Dosimetric leaf gap (DLG) measurement with a Farmer

ionization chamber. The results correspond to sweeping gaps shaped
with the distal leaves in a single layer configuration with gap sizes
from 2 to 20 mm. A DLG of 0.13 mm was found from the linear fit
Figure 10. Interleaf leakage - comparison of calculated and measured extrapolation of the accumulated charge versus sweeping gap size.
profiles of a field with distal MLC stack fully closed, proximal fully
open. The interleaf leakage was modeled so that the integral of the
calculated and measured profiles match (within a defined sub-region).
Profiles were obtained at 10 cm depth, with SSD 90 cm, and
100 MU. Machine MU calibration geometry is 100 MU/Gy at 100 SSD
and 1.3 cm depth.

This document contains confidential, proprietary information of Varian Medical Systems. It may not be copied or reproduced without prior
written permission from Varian Medical Systems.
4
 Technical Perspectives | Eclipse Photon Beam Models for Halcyon

The tongue-and-groove effect is an interplay between the Summary

different interleaf leakage and penumbra effects when the
leaf side exposed to radiation has the step closer or further
away from the source. Measurements performed using a IBA
Dosimetry Blue Phantom2 water tank (90 cm SSD, 10 cm
depth) are shown in Figure 12.
A comprehensive description of the photon beam models in
Eclipse for the Halcyon (RDS) treatment delivery device have
been presented. Representative data from all components of
the model are shown.

Comparison of Eclipse calculations with measurement for

both proximal and distal tongue-and-groove leaf patterns
are shown in Figures 13 and 14.

Figure 12. Water scans for the distal leaves corresponding to superposition of various 1 cm leaf patterns (i.e., even leaves extended,
followed by odd leaves extended from the same bank or from the opposite bank, while proximal leaves were fully retracted). Different
scans correspond to 1) distal leaves, bank A; 2) distal leaves, bank B; and distal leaves interdigitation.

Figure 13. Comparison of calculated and measured profile in the
case corresponding to Figure 12. Calculation resolution is 1 mm.
Notice that the water phantom used in calculation ends roughly
outside [-150, 150] mm interval.
Figure 14. Comparison of calculated and measured profiles as
in Figure 13, but with distal leaves fully retracted, while proximal
ones are closed in even/odd pattern. Calculation resolution is 1 mm.
Notice that the water phantom used in calculation ends roughly
outside [-150, 150] mm interval.

This document contains confidential, proprietary information of Varian Medical Systems. It may not be copied or reproduced without prior
written permission from Varian Medical Systems.
5
 APPENDIX A: RDS BEAM DATA ACQUISITION
HARDWARE AND PARAMETERS
Treatment delivery device parameters for PDDs, an absolute penumbra margin of 70 mm for fields
Beam scans were done in Service Mode, with all servos larger than 10 cm and 50 mm for the rest of the fields, and a
turned on and using the highest dose rate of 800 MU/minute stabilization period of 2 seconds.
based on the machine calibration depth of Dmax.
Measurements
Data acquisition hardware and settings
Measurement Crossline Output
Water phantom: PPD Diagonal
Type Profiles Factor
IBA Dosimetry Blue Phantom2
Field Size 2*, 4*, 6, 8, 2, 4, 28 1, 2, 4,
Software: [cm] 10, 20, 28 6, 8, 6, 8, 10,
OmniPro-Accept 7.5 * diode 10, 14, 20,
measurement 20, 24, 28
Scanning detector (offset): 28
IBA Dosimetry CC13 (-1.8 mm) and Sun Nuclear Corporation Fixed Depth 1.3, 5, 10, 20, N/A 1.3, 5, 10, 5
EDGE Detector (+0.3 mm) [cm] 30 20, 30

Reference detector: SSD [cm] 90 90 90 95

IBA Dosimetry CC13
Table 1. Percent depth dose (PDD) and profiles for fields larger
Bias: than 4 cm were acquired with a CC13 (IBA Dosimetry) ionization
+300 V (CC13), 0 V (EDGE) chamber. The small field, 2 cm and 4 cm, were measured with
the EDGE detector (Sun Nuclear Corporation). An effective point
Settings: of measurement correction (-1.8 mm shift) was applied to all
Positioning speed of 10 mm/second, scanning speed of 3 mm/ measurements performed with the CC13 detector, and (+ 0.3 mm
second, output step width of 1 mm for profiles and 0.5 mm shift) for EDGE Detector.

Results
Output factor table

Field Size X [cm]

1 2 4 6 8 10 14 20 24 28

1 0.766 0.792 0.804 0.808 0.810 0.811 0.813 0.814 0.814 0.814

2 0.816 0.855 0.884 0.892 0.897 0.899 0.902 0.904 0.905 0.905

4 0.829 0.885 0.922 0.939 0.947 0.953 0.958 0.962 0.964 0.964

6 0.834 0.893 0.939 0.958 0.971 0.977 0.985 0.991 0.994 0.994
Field Size Y [cm]

8 0.836 0.898 0.949 0.972 0.982 0.992 1.002 1.010 1.013 1.013

10 0.837 0.899 0.952 0.977 0.991 1.000 1.012 1.021 1.024 1.024

14 0.839 0.902 0.957 0.984 1.001 1.012 1.027 1.038 1.043 1.043

20 0.840 0.904 0.961 0.990 1.008 1.020 1.038 1.051 1.057 1.058

24 0.840 0.905 0.962 0.992 1.011 1.024 1.042 1.057 1.062 1.064

28 0.841 0.905 0.963 0.992 1.011 1.024 1.042 1.058 1.064 1.067

Table 2. Output factors versus field size for 95 cm SSD and 5 cm depth. Fields of 1xY and Xx1 cm2 have been measured using the Sun
Nuclear Corporation EDGE Detector diode. The remaining fields have been measured using the CC13 ionization chamber.

This document contains confidential, proprietary information of Varian Medical Systems. It may not be copied or reproduced without prior
written permission from Varian Medical Systems.
6
 Technical Perspectives | Eclipse Photon Beam Models for Halcyon

Percent depth dose

Figure 1. Percent depth dose curves (PDDs), 90 cm SSD.

Crossline profiles

Figure 2. Crossline profiles at 10 cm depth (90 cm SSD) for square field sizes from 2 to 28 cm.

Diagonal profiles

Figure 3. Diagonal profiles for 28 cm x 28 cm at 5 depths.

This document contains confidential, proprietary information of Varian Medical Systems. It may not be copied or reproduced without prior
written permission from Varian Medical Systems.
1Varian Medical Systems, Palo Alto, California, USA
2Varian Medical Systems, Applied Research, Helsinki, Finland
3Document D29493, Halcyon™ 1.0 Representative Beam Data, Varian Medical Systems
4Document D22358, TPS-AAA-Acuros® XB System Requirement Specifications, Varian Medical Systems

USA, Corporate Headquarters EMEIA and CIS Latin American Asia Pacific Australasian
and Manufacturer Headquarters Headquarters Headquarters Headquarters
Varian Medical Systems, Inc. Varian Medical Systems Varian Medical Systems Varian Medical Systems Varian Medical Systems
3100 Hansen Way International AG Brasil Ltda. Pacific, Inc. Australasia Pty Ltd.
Palo Alto, CA 94304 Cham, Switzerland São Paulo, Brazil Kowloon, Hong Kong Sydney, Australia
Tel. 650.493.4000 | varian.com Tel. 41.41.749.88.44 Tel. 55.11.3457.2655 Tel. 852.2724.2836 Tel. 61.2.9485.0111

Not all features or products are available in all markets.

Intended Use Summary
Varian Medical Systems’ linear accelerators are intended to provide stereotactic radiosurgery and precision radiotherapy for lesions, tumors, and conditions
anywhere in the body where radiation treatment is indicated.
Important Safety Information
Radiation treatments may cause side effects that can vary depending on the part of the body being treated. The most frequent ones are typically temporary
and may include, but are not limited to, irritation to the respiratory, digestive, urinary or reproductive systems, fatigue, nausea, skin irritation, and hair loss. In
some patients, they can be severe. Treatment sessions may vary in complexity and time. Radiation treatment is not appropriate for all cancers.
Medical Advice Disclaimer
Varian as a medical device manufacturer cannot and does not recommend specific treatment approaches. Individual treatment results may vary.

© 2017 Varian Medical Systems, Inc. All rights reserved. Varian and Varian Medical Systems are registered trademarks, and Eclipse and Halcyon are trademarks of Varian
Medical Systems, Inc. All other trademarks are the property of their respective owners.

RAD 10456 09/2017

This document contains confidential, proprietary information of Varian Medical Systems. It may not be copied or reproduced without prior
written permission from Varian Medical Systems.