PHYSICS BRAINLAB · Technical Reference Guide Rev. 2.2 Brainlab Physics 9 1.4 Using the System Purpose and Audience • This guide provides a background and reference for the medical

BRAINLAB PHYSICSRT ELEMENTS BRAINLAB PHYSICS

Technical Reference GuideRevision 2.2Copyright 2019, Brainlab AG Germany. All rights reserved.

TABLE OF CONTENTS1 GENERAL INFORMATION.............................................................................................7

1.1 Contact Data ........................................................................................................................................7

1.2 Legal Information ...............................................................................................................................8

1.3 Symbols................................................................................................................................................9

1.4 Using the System .............................................................................................................................10

1.5 Training and Documentation..........................................................................................................11

2 BASIC INFORMATION.....................................................................................................13

2.1 Safety Notes ......................................................................................................................................13

2.2 Treatment Field Setup .....................................................................................................................172.2.1 Leakage Radiation Caused by Closed MLC Leaf Gaps....................................................................19

2.3 Measurement for Small Radiation Fields ....................................................................................21

2.4 Beam Data Measurement Methods...............................................................................................232.4.1 Raw Data Mode in Physics Administration .......................................................................................242.4.2 Pencil Beam Raw Data Mode...........................................................................................................252.4.3 Entering Machine Profile Data using Brainlab Excel Templates (Optional) .......................................28

3 PENCIL BEAM: ALGORITHM...................................................................................31

3.1 Pencil Beam Dose Algorithm .........................................................................................................313.1.1 Pencil Beam for Dynamic Conformal Arc..........................................................................................39

3.2 Limitations of the Pencil Beam Algorithm ..................................................................................403.2.1 Extrapolation Outside the Range of Measured Values.......................................................................403.2.2 Other Limitations ..............................................................................................................................41

4 PENCIL BEAM: GENERAL BEAM DATA MEASUREMENT .........43

4.1 Introduction .......................................................................................................................................434.1.1 Recommended Equipment ...............................................................................................................454.1.2 General Measurement Requirements...............................................................................................46

4.2 Absolute Linac Calibration.............................................................................................................47

4.3 Background Leakage.......................................................................................................................50

4.4 Depth Dose Profile ...........................................................................................................................51

TABLE OF CONTENTS

Technical Reference Guide Rev. 2.2 Brainlab Physics 3

4.5 Scatter Factors (Output Factors) ..................................................................................................53

4.6 Diagonal Radial Profiles..................................................................................................................55

4.7 Transversal Profiles .........................................................................................................................584.7.1 Measurement Using a Water Phantom and High-Resolution Detector..............................................604.7.2 Film Dosimetry Measurement...........................................................................................................61

4.8 Dynamic Leaf Shift Measurements...............................................................................................62

4.9 Verification of Radiologic Field Corrections ..............................................................................63

5 PENCIL BEAM: BEAM DATA CHECKLISTS ..............................................65

5.1 Beam Data for Brainlab m3 ............................................................................................................655.1.1 Additional Information.......................................................................................................................675.1.2 Transversal Profile Shape ................................................................................................................68

5.2 Beam Data for Elekta Agility ..........................................................................................................695.2.1 Additional Information.......................................................................................................................715.2.2 Transversal Profile Shape ................................................................................................................72

5.3 Beam Data for MHI MLC 60.............................................................................................................735.3.1 Additional Information.......................................................................................................................745.3.2 Transversal Profile Shape ................................................................................................................75

5.4 Beam Data for Novalis.....................................................................................................................765.4.1 Additional Information.......................................................................................................................785.4.2 Transversal Profile Shape ................................................................................................................79

5.5 Beam Data for Varian HD120 (SRS Flattening Filter) ................................................................805.5.1 Additional Information.......................................................................................................................825.5.2 Transversal Profile Shape ................................................................................................................83

5.6 Beam Data for Varian HD120 (Standard Irradiation Mode and Flattening Filter FreeMode) ..........................................................................................................................................................84

5.6.1 Additional Information.......................................................................................................................865.6.2 Transversal Profile Shape ................................................................................................................87

5.7 Beam Data for Varian 120 (SRS Flattening Filter)......................................................................885.7.1 Additional Information.......................................................................................................................905.7.2 Transversal Profile Shape ................................................................................................................91

5.8 Beam Data for Varian 120 (Standard Irradiation Mode and Flattening Filter Free Mode) ...........................................................................................................................................................................92

5.8.1 Additional Information.......................................................................................................................945.8.2 Transversal Profile Shape ................................................................................................................95

6 MONTE CARLO: ALGORITHM................................................................................97

6.1 Introduction to the Monte Carlo Algorithm.................................................................................976.1.1 Brainlab Monte Carlo Algorithm........................................................................................................98

6.2 The Virtual Energy Fluence Model (VEFM) .................................................................................99

TABLE OF CONTENTS

4 Technical Reference Guide Rev. 2.2 Brainlab Physics

6.3 Modeling of the Collimating System ..........................................................................................101

6.4 The MC Patient Dose Computation Engine ..............................................................................103

6.5 MC Parameters................................................................................................................................105

7 MONTE CARLO: GENERAL BEAM DATA MEASUREMENT....109

7.1 Introduction .....................................................................................................................................1097.1.1 Recommended Equipment ............................................................................................................. 111

7.2 Coordinate Systems ......................................................................................................................112

7.3 Data Correction ...............................................................................................................................113

7.4 Beam Data Measurements in Air.................................................................................................114

7.5 Beam Data Measurements in Water ...........................................................................................116

8 MONTE CARLO: BEAM DATA CHECKLISTS .........................................117

8.1 Beam Data for Elekta Agility ........................................................................................................117

8.2 Beam Data for MHI MLC 60...........................................................................................................120

8.3 Beam Data for Novalis/Brainlab m3 ...........................................................................................122

8.4 Beam Data for Varian HD120 (SRS Flattening Filter) ..............................................................124

8.5 Beam Data for Varian HD120 (Standard Irradiation and Flattening Filter Free Mode).....126

8.6 Beam Data for Varian 120 (SRS Flattening Filter)....................................................................128

8.7 Beam Data for Varian 120 (Standard Irradiation and Flattening Filter Free Mode) ..........130

9 DYNAMIC DELIVERY .....................................................................................................133

9.1 Introduction .....................................................................................................................................133

9.2 Deliverability of Arcs .....................................................................................................................134

9.3 Leaf Tolerance .................................................................................................................................136

9.4 Dynamic Leaf Shift for Modulated Treatments.........................................................................137

10 QUALITY ASSURANCE ............................................................................................139

10.1 Introduction to Quality Assurance ...........................................................................................13910.1.1 Required Equipment.....................................................................................................................140

10.2 Machine-Related Quality Assurance ........................................................................................14110.2.1 Specific Tests ...............................................................................................................................142

10.3 Patient-Related Quality Assurance ..........................................................................................14410.3.1 Recommended Procedures..........................................................................................................145

TABLE OF CONTENTS


10.4 Patient-Specific Quality Assurance..........................................................................................14610.4.1 Pre-Treatment Patient QA ............................................................................................................14710.4.2 General Patient QA ......................................................................................................................148

11 APPENDIX 1 .........................................................................................................................149

11.1 Accuracy of Dose Algorithms....................................................................................................14911.1.1 Pencil Beam and Monte Carlo .......................................................................................................149

11.2 Limitations of Dose Algorithms ................................................................................................151

12 APPENDIX 2.........................................................................................................................153

12.1 Linac Energy .................................................................................................................................153

13 APPENDIX 3.........................................................................................................................155

13.1 Bibliography ..................................................................................................................................155

TABLE OF CONTENTS


1 GENERAL INFORMATION1.1 Contact Data

Support

If you cannot find information you need in this guide, or if you have questions or problems, contactBrainlab support:

Region Telephone and Fax Email

United States, Canada, Centraland South America

Tel: +1 800 597 5911Fax: +1 708 409 1619

[email protected]

Brazil Tel: (0800) 892 1217 [email protected]

UK Tel: +44 1223 755 333

[email protected]

Spain Tel: +34 900 649 115

France and French-speakingregions Tel: +33 800 676 030

Africa, Asia, Australia, EuropeTel: +49 89 991568 1044Fax: +49 89 991568 811

JapanTel: +81 3 3769 6900Fax: +81 3 3769 6901

Feedback

Despite careful review, this user guide may contain errors. Please contact us [email protected] if you have improvement suggestions.

Manufacturer

Brainlab AGOlof-Palme-Str. 981829 MunichGermany

GENERAL INFORMATION


mailto://[email protected]



mailto:[email protected]

1.2 Legal Information

Copyright

This guide contains proprietary information protected by copyright. No part of this guide may bereproduced or translated without express written permission of Brainlab.

Brainlab Trademarks

• Brainlab® is a trademark of Brainlab AG in Germany and/or the US.

Non-Brainlab Trademarks

• Dosimetry-PRO® is a registered trademark of VIDAR Systems Corporation.• Kodak® is a registered trademark of Eastman Kodak Company.• Microsoft® and Windows® are registered trademarks of Microsoft Corporation.

Patent Information

This product may be covered by one or more patents or pending patent applications. For details,see: www.brainlab.com/patent.

CE Label

• The CE label indicates that the Brainlab product complies with the essential re-quirements of European Council Directive 93/42/EEC, the Medical Device Di-rective (MDD).

• According to the rules established by the MDD, the classification of the Brain-lab product is defined in the corresponding Software User Guide.

Sales in US

US federal law restricts this device to sale by or on the order of a physician.

Legal Information


https://www.brainlab.com/patent/

1.3 Symbols

Warnings

WarningWarnings are indicated by triangular warning symbols. They contain safety-criticalinformation regarding possible injury, death or other serious consequences associatedwith device use or misuse.

Cautions

Cautions are indicated by circular caution symbols. They contain important informationregarding potential device malfunctions, device failure, damage to device or damage toproperty.

Notes

NOTE: Notes are formatted in italic type and indicate additional useful hints.

GENERAL INFORMATION


1.4 Using the System

Purpose and Audience

• This guide provides a background and reference for the medical physics required to correctlyoperate Brainlab’s radiotherapy treatment planning software.

• This guide is written for all members of the clinical team who use or handle Brainlabradiotherapy treatment planning software, in particular medical physicists.

• You should read this guide carefully and acquaint yourself sufficiently with the software beforeuse.

Operator Profile

WarningBrainlab planning software and accessory devices may only be operated by qualifiedmedical professionals.

Plausibility Review

WarningAll information input to the Brainlab planning application and all information received fromthe Brainlab planning application as output must be reviewed regarding its plausibilitybefore patient treatment.

Compatibility

WarningOnly medical devices and spare parts specified by Brainlab may be used with Brainlabplanning software. Using unauthorized devices or spare parts may adversely affect thesafety and/or effectiveness of the Brainlab planning software and endanger the safety ofthe patient, user and/or environment.

Available Functions

This guide contains information on various algorithms and supported hardware functionality.Depending on the license you purchased, software version, and national regulatory requirements,some of these algorithms or functionality may not be available.

Further Information

For specific information on the intended use of Brainlab’s radiotherapy treatment planningsoftware, and on related compatibilities, refer to the appropriate Brainlab RT Elements SoftwareUser Guide.

Using the System


1.5 Training and Documentation

Brainlab Training

In order to ensure safe and appropriate use of the system, Brainlab recommends that beforeusing a Brainlab planning application for the first time, all users should participate in an extensivetraining program held by a Brainlab representative.

Responsibility

WarningBrainlab planning applications are solely designed to provide additional assistance tomedical staff. They do not substitute or replace user experience, or invalidate userresponsibility during their use.

WarningEnsure that individuals authorized to perform treatment planning functions areappropriately trained for the function they perform.

Reading User Guides

Successful and safe treatment using Brainlab planning software requires careful proceduralplanning.It is therefore important that all users of the software:• Read the relevant user guides carefully before using the software• Have access to these user guides at all times

Available User Guides

User Guide Contents

Physics AdministrationSoftware User Guide

Detailed instructions on using the Physics Administration appli-cation.

Software User Guides Detailed instructions on using Brainlab Elements.

GENERAL INFORMATION


Training and Documentation


2 BASIC INFORMATION2.1 Safety Notes

Important Notes on System Safety

This section contains important information that must be considered for the safe and effectiveoperation of the treatment planning system.Refer to the appropriate Brainlab Elements software user guide and the Physics AdministrationSoftware User Guide.

WarningIt is your responsibility to establish a comprehensive quality assurance program suitablefor detecting errors, limitations or inaccuracies of the treatment planning and treatmentdelivery systems. For more details, consult the quality assurance chapter in this technicalreference guide.

WarningMeasure the absolute accuracy of the Brainlab treatment planning system in combinationwith the used treatment delivery systems using phantoms. The measured accuracy mustbe taken into account when configuring plan parameters in order to ensure accuratetreatment delivery.

WarningEnsure proper delivery of the treatment plan to the patient. It is strongly recommended toperform a phantom verification for every treatment plan using exactly the same parametersettings that will be used for the real patient during actual treatment.

WarningTo ensure consistency of the positioning and the treatment parameters in case severalplanning systems are used, all final positioning and treatment parameters must beprovided by the same treatment planning system. For example, if information from theBrainlab RT Elements software is used directly for patient positioning then correspondingtreatment plan must come from the same Brainlab RT Elements software and the treatmentplan must not be modified with any other treatment planning system.

WarningStereotactic treatments, such as stereotactic radiosurgery (SRS), incorporate very highdose rates and doses per fraction and are typically planned with reduced target volumemargins. Therefore, you must use additional safety precautions during treatment planning,plan transfer and treatment delivery. It is highly recommended to perform additional qualityassurance before each stereotactic patient treatment.

WarningIn general, it is not the intended use of the system to treat a patient using a plan withprimary jaws blocking the MLC aperture. The only exception is the Elekta Agility MLCbecause of the behavior of the guard leaves and the Jaw Tracking feature. In treatmentplans for the Elekta Agility MLC, the jaws overlap the MLC field. To use the Elekta AgilityMLC, additional beam data measurements specified in the Brainlab Physics TechnicalReference Guide are necessary.

BASIC INFORMATION


WarningAll treatment plan reports must be approved by a qualified person before the informationthey contain is used for radiotherapy treatment purposes.

WarningBrainlab recommends using the treatment plan reports to verify all treatment parametersincluding, but not limited to collimator sizes and positions, device angles and dosespecifications, directly at the treatment delivery system.

WarningBefore starting patient treatment, you must first complete system acceptance, verificationand validation of the treatment planning system, including the machine profiles.

WarningFor every patient treatment plan, you must verify that the linac configuration planned wascorrectly transferred and applied at the linac. This includes, but is not limited to, theflattening filter mode and accessory configurations.

WarningAlways be aware that the quality of the output depends critically on the quality of the inputdata. Any irregularities or uncertainties about input data units, identification, or qualityissues of any other nature must be thoroughly investigated before the data is used.

WarningEnsure that your imaging devices (e.g., CT scanner) are properly configured andcalibrated. Regularly check the calibration by imaging and verifying test phantoms.

WarningCheck the accuracy of the resulting outer contour and tissue model used for dosecalculations. The entirety of the area relevant for treatment must be included within thecontour.

WarningEnsure during the complete planning process that you are working on the correct patientdata set. The patient information is displayed in the main screen of the Brainlab treatmentplanning system.

WarningAlways make sure that the treatment delivery systems used for treating a patient are thesame as intended during the planning process, i.e., the selected machine profilecorresponds to the treatment machine.

WarningEnsure that any treatment accessories, such as the conical collimator mount and conicalcollimator of the size which is required by the treatment plan, are installed properly in thebeam path before delivery. Your treatment delivery systems may not be able to assert abeam interlock in case of missing or wrong accessories. Refer also to the documentationprovided for the accessories.

WarningPrior to treatment, it is your responsibility to verify from inside the treatment room that theselected gantry and table angles can be used to perform the treatment without resulting ininjury to patients or damage to equipment such as the treatment delivery system.

WarningIt is recommended to verify and confirm the patient setup with an appropriate positionverification method before treatment. Sample plans may be generated using phantoms totest the accuracy of the actual patient setup method.

Safety Notes


WarningEnsure that the Brainlab treatment planning system is correctly configured and that theconfiguration reflects the parameters of the treatment delivery systems. These parametersinclude, but are not limited to, the linac scale convention, mechanical limits or dosimetricparameters such as linac energy or fluence modes like SRS and FFF (flattening-filter-free).

WarningIt is your responsibility to ensure that the machine configurations are synchronizedbetween the different configuration locations (e.g., treatment planning system, record andverify system and treatment delivery system) at any time. A mismatch in the machineconfiguration used for planning and the one used for treatment can lead to unintendedtreatment delivery or a disruption in the clinical workflow.

WarningBrainlab provides up-to-date measurement instructions. Ensure that the latestmeasurement instructions are used during beam data acquisition. For more informationcontact your Brainlab support specialist.

WarningMake sure that your beam data measurements are up-to-date and that dose algorithms areproperly configured and calibrated. Regularly check the configuration and calibrationusing phantom measurements.

WarningIf one or more components of the treatment delivery system have been modified,exchanged or recalibrated, a revalidation of the treatment planning system in combinationwith the treatment delivery system must be performed in accordance with your qualityassurance procedures. If components have been modified that influence the dosimetricparameters of the system, the beam data measurements must be repeated and the reviseddata entered into the system using Physics Administration.

WarningPre-installed drivers, etc. should only be changed if absolutely necessary. In case of adriver update, a virus scanner update or similar, ensure that the Brainlab treatmentplanning system performs unchanged. A software revalidation is strongly recommended.

WarningYou must carefully obey the specifications and recommendations given by themanufacturer of your dosimetry equipment. Especially all dose detectors have a clearlyspecified range of field sizes they are applicable for. Using dosimetry equipment out of itsspecifications or in a wrong way may lead to wrong dose calculations.

WarningThe Brainlab beam data (e.g. Novalis Tx) installed during system acceptance is for testpurposes only and is not suitable for clinical use.

WarningWhen measured data is sent to Brainlab, the following applies:• Brainlab has no possibility to verify the correctness of any data received from or

returned to a user.:• Any feedback or recommendations from Brainlab are based on the data received and

depend on the correctness of the data itself.• If received data has been processed by Brainlab and returned to you, it is not ensured

that the returned data is correct.You are fully responsible for:- Verifying the correctness of the data received from Brainlab- Verifying the correctness of any feedback or recommendations provided by Brainlab- Validating the safety and effectiveness of the data returned by Brainlab before

performing any patient treatment

BASIC INFORMATION


• The fact that Brainlab may have processed certain data, does not affect your overallresponsibility to check the correctness of the final beam profile.

Safety Notes


2.2 Treatment Field Setup

Skin Dose Build-Up

If irradiation is directed through a solid carbon fiber layer, attenuation and dose build-up occur. • These effects can be observed in all sandwich design couch tops, similarly designed couch

inserts and immobilization devices with solid carbon fiber panels.• They are due to the high-density properties of carbon fiber, and can vary (e.g., depending on

the beam energy and the entry angle to the couch top).

Comparison of Dose Build-Up

In order to illustrate the effect of couch top usage on patient dose, the following examplecompares the dose build-up and dose attenuation in a water phantom when treatment isperformed with the Brainlab Imaging Couch Top, and when no couch top is used.Comparison of dose build-up at 6 MV (Monte Carlo Calculation):

①

②

Figure 1

Explanation of the Dose Build-Up Diagram

In the figure above, the couch top begins at 0.1 cm with a 0.2 cm carbon fiber layer. This isfollowed by 4.5 cm of plastic foam and then by another 0.2 cm of carbon fiber. The water layerrepresenting the patient begins at a depth of 5 cm.• The red curve ① shows the depth dose for a given number of monitor units in the water

phantom when no couch top is used.• The blue curve ② shows the depth dose for the same number of monitor units in the water

phantom when the Brainlab Imaging Couch Top is used.• Both curves are calculated with a Monte Carlo algorithm for a 6 MV photon beam.

Note that the blue curve ② indicates an increase in dose (skin dose) when the beam reaches thewater phantom representing the patient. This is due to the large number of secondary electronsgenerated as a result of the relatively high density of the carbon fiber material. These electrons hitthe skin surface, resulting in the high dosage shown.

BASIC INFORMATION


Verification

As most radiotherapy treatment planning software assumes that only air is present between thelinac and the skin surface, they are unable to model effects of this kind or take them intoconsideration. These effects must therefore be verified experimentally with appropriatemeasurements, for example using a phantom with PMMA slabs. The attenuation and dose build-up results obtained from such experiments must then be taken into account during treatmentplanning.In Brainlab Elements, a couch top can be selected during surface segmentation so that effectssuch as attenuation and dose build-up can be taken into account during treatment planning.

WarningTreating patients through the couch top from a posterior angle should be avoided. If itunavoidable, you should include your couch top model for dose calculation. If not, beaware that the calculated dose distribution does not include the additional attenuation orthe increased skin dose close to the couch top. The dose calculation must be correctedmanually for these factors.

Safety Notes

WarningInaccurate determination of the skin surface due to poor or incomplete image data canresult in incorrect calculation of the equivalent depth. The entry region of each beam or arcmust therefore be verified in your planning software using the depth view provided.

Treatment Field Setup


2.2.1 Leakage Radiation Caused by Closed MLC Leaf Gaps

Background

Treatment plans for MLCs typically contain closed leaf pairs. Ideally, no dose is delivered throughthe small gap remaining between the closed leaf tips. However, depending on the leaf tip designof the MLC, a certain leakage is technically unavoidable unless this leaf gap is covered by thelinac jaws or any other additional collimating device.

Illustration

The diagram shows a closed MLC leaf pair where the leaf gap is not covered as opposed to a paircovered by linac jaws:

①

②

③④

Figure 2

No. Component

① Position of linac jaws

② Closed mMLC leaf pair

③ Resulting dose

④ Leakage dose

To avoid delivery of undesired leakage dose to the patient, it is essential that the gap betweenclosed leaf pairs is always completely shielded by the linac jaws.For some linac MLC combinations it is technically impossible to shield the closed leaf gaps.However, since the dose algorithms are able to predict the leakage radiation caused by the closedleaf gaps, it is possible to verify the influence of the additional radiation on the patient treatment.

Details

There is a small area with less radiation shielding between the opposing leaf tips of closed leafpairs. This results from the technical design of the MLC, i.e. the shape of the leaf's tip and a smallgap remaining between the leaves.If this leaf gap is not covered by e.g., the linac jaws, leakage radiation can pass through this gap.The amount of this leakage dose depends on the dose delivery system and mainly on theindividual treatment plan. Compared to the planned treatment dose, plans with complex leafsequences could in particular result in a significant leakage dose.

BASIC INFORMATION


The leakage dose for a treatment plan can be determined by phantom measurements usingappropriate equipment, such as radiation sensitive films.Bear in mind that linac jaws also are subject to mechanical positioning uncertainties. These linacjaw positioning accuracy limitations must be considered when defining jaw positions for the use ofthe MLC. For details, consult the user guide and specifications of your linac.Due to the hardware limitations of some MLCs (such as the Elekta Agility) the leaves may notalways conform to the target region and the required static leaf gaps are not completely coveredby primary jaws.

WarningIt is technically not always possible to move closed leaf pairs behind jaws. Double-checkwhether the closed leaf pairs have been positioned behind jaws. If this is not the case, it isyour responsibility to decide whether the dose leakage due to this is acceptable or not.

How to Verify Your System

Brainlab RT Elements provide the functionality to automatically place the leaf gap of the closedleaves behind the linac jaws during treatment planning. Follow the steps summarized below toverify that your system is setup correctly.

Step

1. Perform a suitable measurement to determine the maximum linac jaw positions that stillcompletely cover the MLC leaf gap if the leaves are closed at the maximum distance fromthe central beam axis.

2. Check your machine profile / beam profile using Physics Administration to:• Make sure that the linac jaw motion limits are smaller than or equal to the maximum

jaw positions determined in step 1.• Make sure that the leaf gap of the closed leaves is automatically positioned behind the

linac jaws.For this check, or to adjust the linac jaw motion limits to the adequate values, follow theinstructions described in the Physics Administration Software User Guide.

Leakage Radiation Caused by Closed MLC Leaf Gaps


2.3 Measurement for Small Radiation Fields

General Recommendations

Specific measurements must be completed before performing stereotactic treatments with verysmall field sizes. These measurements must be based on valid international dosimetry standardsfor small fields, especially IAEA TRS-483 (2017). Report IAEA TRS-483 contains an internationalcode of practice for reference and relative dose determination for small static fields used inexternal beam radiotherapy.Assuming that averaging of the inhomogeneous dose within the sensitive detector volume resultsin a reduced detector signal, higher values from smaller detectors are likely to be closer to the truevalue. For this reason, the smallest detector available should be used when performing small fielddosimetry (Alfonso et all 2008 and Sauer et al 2007). For central axis measurements, such asdepth dose, tissue phantom ratios and scatter/output factors, the detector dimensions should besignificantly smaller than the field size.

Code of Practice

Special care is required when selecting and handling the required dosimetry equipment. For smallfield sizes, it is particularly important to correctly align the water phantom and the detectormovement direction in relation to the beam axis and the beam center (refer to e.g., IAEATRS-483). Even if the detector size is suitable for the small fields to be measured, accuratesensitivity corrections (e.g., energy dependency of the detector signal or fluence perturbationeffects) must be applied in accordance with the specifications provided by the detectormanufacturer.For many different detector types from a variety of vendors, tables 23 to 27 of IAEA TRS-483provide field output correction factors. These correction factors (if available) shall be appliedduring scatter/output factor determination. Before sending the measurement data to Brainlab,indicate either in the Excel template or using the Raw Data mode of Physics Administration ifthe scatter/output factors are corrected according to IAEA TRS-483 or not. For instructions, referto the Excel template and to the Physics Administration Software User Guide.When correcting scatter/output factors according to IAEA TRS-483, please also take into accountthe uncertainties of the correction factors provided by the report (see table 37 of IAEA TRS-483).These uncertainties are different for the different groups of detectors and they are dependent onthe field size.For more details, refer to the respective publications (e.g., Das, et al. 2008, IPEM Report Number103 2010 or Wuerfel 2013). Whenever possible, follow the code of practice as provided by reportIAEA TRS-483.

Ensuring Accuracy

When treating small field sizes, the dose profile will either show only a narrow plateau region or noplateau at all. If the sensitive volume of the detector is too large, the measured dose will be lowerthan the real dose, resulting in radiation overdose. The use of incorrectly sized sensitive volumesis a major contributing factor in inaccurate dose measurement.

WarningCarefully obey the specifications and recommendations provided by the manufacturer ofyour dosimetry equipment. Dose detectors in particular have a clearly specified range offield sizes for which they may be used. Using a dose detector for an application for which itwas not intended, or in the wrong orientation, may lead to incorrect dose calculations.

WarningThe measurement of dose for small radiation fields (less than 30x30 mm2 field size) has tobe done using equipment that is suitable for these field sizes.

BASIC INFORMATION


WarningFor MLCs with a relatively large required minimum leaf gap, do not plan treatments for verysmall or narrow targets.

Measurement for Small Radiation Fields


2.4 Beam Data Measurement Methods

Measuring Beam Data

You can measure beam data in these ways:

Method See

Pencil BeamRaw Data mode of Physics Administration Page 25

Excel Template method (optional) Page 28

Monte Carlo Raw Data mode of Physics Administration Page 113

BASIC INFORMATION


2.4.1 Raw Data Mode in Physics Administration

Background

The Raw Data mode of Physics Administration enables you to enter measured beam data priorto processing. The Raw Data mode no longer requires you to collect the data in Excel templates.Using Physics Administration, raw data can be converted to beam profiles, which then can beused with machine profiles for treatment planning.There are Raw Data modes for Pencil Beam and Monte Carlo measurement data. For details ofhow to use the Raw Data modes, refer to the Physics Administration Software User Guide.For the Pencil Beam of RT Elements, beam data collection and beam data processing using theRaw Data mode of Physics Administration is recommended. The Excel template method (seepage 28) is optional.It is not possible to mix both methods. For a certain MLC and energy, all Pencil Beam data needto either be collected using the Raw Data method or the Excel method.

Raw Data Mode in Physics Administration


2.4.2 Pencil Beam Raw Data Mode

Background

The following are different in the Pencil Beam Raw Data mode compared to the Excel templateapproach:• Depth dose profiles can have individual depth-coordinate values for each field size.• Diagonal profiles can have individual radius-coordinate values for each depth.• Diagonal profiles must be processed with Physics Administration to get the Radial Factors.• Transversal profiles can have individual coordinate values for each depth.• Transversal profiles must be processed with Physics Administration to determine Source

Function Correction and radiological shift parameters.

Scatter Factors (Output Factors)

The measurement instruction setup and workflow are described on page 53. Before enteringmeasurement results, enter Source Surface Distance (SSD) and measurement depth of thescatter measurement in the Scatter Data dialog.These values must be the same as the SSD and measurement depth of the Nominal Linac Outputmeasurement. Otherwise, generation of a beam profile is impossible.If needed, adjust the MLC and jaw size values in the scatter table section of the dialog and enteryour scatter data.• The gray fields in the sample matrix tables provided on page 65 must be measured in all

cases.• The white fields represent MLC and jaw combinations that are not recommended for use with

Brainlab’s radiotherapy treatment planning software.• It is therefore not necessary to measure these larger MLC fields. Instead, it is sufficient to enter

the last mandatory value measured, e.g. in the case of a jaw setting of 60 x 60 mm² you canuse the value measured for the 60 x 60 mm² MLC field (0.8710 on page 86).

You can also paste an entire scatter table at once using the paste button. In this case, the MLCand jaw sizes are automatically adjusted. For more details, refer to the Physics AdministrationSoftware User Guide.

Dynamic Leaf Shift

The measurement instruction setup and workflow are described on page 62.Enter the results of the dynamic leaf shift measurement in the Dynamic Leaf Shift dialog forcalculation.

Nominal Linac Output and Background Leakage

The measurement instruction setup and workflow are described on page 47 and page 50.The Nominal Linac Output data must be entered in the Nominal Linac Output dialog of the PencilBeam Raw Data interface:

Step

1. Enter your absolute linac calibration by defining the Source Surface Distance,Measure-ment Depth, Normalization Field Size and Nominal Linac Output.

2. Enter Leakage for Open Jaws and Leakage for Closed Jaws in the Multileaf Back-ground Leakage section.

Depth Dose Profile

The measurement instruction setup and workflow are described in page 51. Depth dose profilescan be measured in a PDD-like (fixed SSD) or TPR-like (isocentric) setup.

BASIC INFORMATION


To enter measurement results:

Step

1. Define the measurement setup (fixed SSD or isocentric) in the Depth Dose dialog. If afixed SSD approach has been used, you also need to enter the SSD of the PDD-meas-urement in the dialog.

2. If needed, adjust the depth dose field sizes using the Add and Remove buttons in thecontrol area.

3. Paste each depth dose profile in the corresponding Depth Dose Profile dialog. Differentcoordinate values may be used for each field size.

4. You can also paste a table with depth dose data for several field sizes at once usingPaste Profiles. In this case, field sizes are automatically adjusted. For more details, referto the Physics Administration Software User Guide.

5. Depth dose profiles can be normalized arbitrarily. However, normalization to a commonreference depth or maximum may simplify consistency checks.

Diagonal Radial Profiles

The general measurement instruction setup and workflow are described in page 55. Diagonalprofiles can be measured in a PDD-like (fixed SSD) or TPR-like (isocentric) setup.NOTE: To get diagonal profiles suitable for Radial Factor calculation, add-on MLCs (e.g. Brainlabm3) must not be detached for the diagonal profile measurement.

The MLC leaves must be retracted.To enter measurement results:

Step

1. Define the measurement setup (fixed SSD or isocentric) in the Diagonal Profiles dialog.If a fixed SSD approach has been used, you also need to enter the SSD of the measure-ment in the dialog.

2. If needed, adjust the depths using the Add and Remove buttons in the control area.

3. Paste each diagonal profile in the corresponding Diagonal Profile Data dialog. Differentcoordinate values may be used for each depth.

4. You can also paste a table with diagonal profile for several depths at once using PasteProfiles. In this case, depths are automatically adjusted. For more details, refer to thePhysics Administration Software User Guide.

5. Diagonal profiles can be normalized arbitrarily.

After entering all diagonal profiles (and depth dose data), Radial Factors can be calculated. Referto the Physics Administration Software User Guide for more details.

Transversal Profiles

The measurement instruction setup and workflow are described on page 60 and page 61.To enter measurement results:

Step

1. Define the measurement setup (fixed SSD or isocentric) in the Transversal Profiles dia-log. If a fixed SSD approach has been used, you must also enter the SSD of the meas-urement in the dialog.

2. If needed, adjust the depths using the Add and Remove buttons in the control area.

3. Paste each transversal profile in the corresponding Transversal Profile Data dialog. Dif-ferent coordinate values may be used for each depth and direction.

Pencil Beam Raw Data Mode


Step

4. You can also paste a table with transversal profiles for several depths at once usingPaste Profiles. In this case, depths are automatically adjusted. For more details, refer tothe Physics Administration Software User Guide.

5. Transversal profiles can be normalized arbitrarily.

After entering all transversal profiles (and after calculating the Radial Factors), Source FunctionCorrection and radiologic leaf shift parameters can be calculated. Refer to the PhysicsAdministration Software User Guide for more details.

BASIC INFORMATION


2.4.3 Entering Machine Profile Data using Brainlab Excel Templates (Optional)

Excel Method Restrictions

In contrast to the Raw Data method, the Excel method has the following restrictions:• Depth dose profiles need to use the same depth-coordinates for all field sizes• Diagonal profiles should use the same radius-coordinates for all depths• Transversal profiles should use the same coordinates for all depths• Diagonal and transversal profiles must be processed by Brainlab

NOTE: It is not possible to mix both methods. For a certain MLC and energy, all data need toeither be collected using the Raw Data method or the Excel method.

Transferring Data

When using the Brainlab Excel Templates, there is no direct way to copy the data to PhysicsAdministration (Machine Profile mode). Moving the beam data (e.g. PDD/TMR, Scatter or RadialFactors returned from Brainlab Physics after processing) from the Excel template to PhysicsAdministration (Machine Profile mode) requires a few intermediate steps.The data should be copied to a new Excel workbook and then saved as a tab delimited text file.Then this data must to be transferred to the Brainlab workstation (e.g. through a mapped drive ora USB drive), where it can be opened as a text file and then copied and pasted into the machineprofile. This needs to be done for each table containing beam data (e.g. PDD/TMR, Scatter orRadial Factors returned from Brainlab Physics after processing).

Workflow

Figure 3

Step

1. Copy data into a new Excel workbook:• Select the entire table• Paste data into a new Excel Workbook

2.

Save the workbook as a text file type (Tab delimited) (*.txt).

3. Transfer the tab delimited text file to the Brainlab workstation.

4. Open the text file on the Brainlab workstation using the text editor Notepad.

Entering Machine Profile Data using Brainlab Excel Templates (Optional)


Step

5. Select all the data in the text file and copy.

6. Paste the data into the appropriate table in Physics Administration:

• Select the 'empty box' in the top left of the table.

• Select Paste from the Edit drop down menu (or use the keystroke combination Ctrl+V).

BASIC INFORMATION


Entering Machine Profile Data using Brainlab Excel Templates (Optional)


3 PENCIL BEAM:ALGORITHM

3.1 Pencil Beam Dose Algorithm

Background

Pencil Beam algorithms are well established and accepted methods to calculate dose distributionsin radiotherapy.In the Brainlab Pencil Beam dose algorithm the incident beams are divided into many thinbeamlets. For each beamlet, an individual radiological path length correction is performed tocorrect for tissue density inhomogeneities, taking into account even very small-structuredinhomogeneities. The Fast Fourier Transformation (FFT) is used for the beam-kernel convolutionwith the fluency distribution of the beam. The algorithm uses fast ray tracing and adaptive gridcalculations that reduce the number of necessary dose computations. Due to these optimizations,2D and 3D dose distributions can be calculated within milliseconds.The Brainlab Pencil Beam dose algorithm is based on publications by Mohan et al (1985, 1986,and 1987). It is implemented to work for conformal beam, conformal arc, IMRT and VMATtreatments. This chapter describes the dose algorithm as it is applied in Brainlab planningsoftware for the different treatment modules.

Monoenergetic Pencil Beam (MPB)

In the following, the definition monoenergetic Pencil Beam is used for a parallel monoenergeticphoton beam with energy E and an infinitesimal cross section. A Pencil Beam incident on ahomogeneous water phantom perpendicular to the surface gives rise to a dose distribution.Assuming a linear attenuation of the photon fluence in water, the number of first collisions in a unitvolume taking place at a depth d below the water surface is given by

NOF E( ) eμwater E( ) d⋅–

μwater E( )⋅ ⋅where:

NOF E( ) Number of photons with energy E, averaged over open field. The radial var-iation of the beam intensity is incorporated at a later point.

μwater E( ) Linear attenuation coefficient of photons in water.

d Depth of observation point in water.

Differential Pencil Beam (DPB)

A differential Pencil Beam (DPB) describes the dose distribution relative to the first collision of amonoenergetic Pencil Beam of photons in an infinitely large homogeneous medium. The dosedistribution kPB, diff (E, lPQ, θPQ) is a function of the photon energy E, the distance lPQ between thepoint of first collision P and the observation point Q and the polar angle θ between the incidentPencil Beam and the scattering direction (see the figure below). The calculation of DPB dose

PENCIL BEAM: ALGORITHM


distributions is performed by a Monte Carlo code for various photon energies between 100 keVand 50 MeV.It takes into account the scattering of secondary photons and electrons up to a certain cutoffenergy.The definition of differential Pencil Beam:

Pencil Beam

PPoint of first collision

(0, 0, 0)

dp

lPQ

QkPB,diff(lPQ, θPQ, E)

θ

Figure 4

Pencil Beam Kernel

Given the DPBs and the number of photons at depth d, the dose of a monoenergetic Pencil Beamat point Q is given by the line integral in the semi-infinite water phantom. Given the acceleratordependent energy distribution NOF(E) of the photon beam, the integration over all energies can beperformed, giving the polyenergetic Pencil Beam kernel.

lPQ Q P–=

θPQ PQ P,( )=

P O O dp, ,( )=

Q x y d, ,( )=

x y d, ,( ) NOF E( ) eμwater E( ) dp⋅–

μwater E( )

kPB diff, E lPQ θPQ, ,( )

⋅⋅ ⋅

E ddp⋅d

=P

Pencil Beam Dose Algorithm


Source Function Correction

An optional source function correction can be applied, which describes the influence of the finitesize of the source, collimator and flattening filter scatter, curvature of leaf ends and other effectsbroadening the penumbra.The source function correction is specified to have a Gaussian distribution with the width sigmaand the amplitude at a certain depth. The width sigma and the amplitude can be specified for twodepths in Physics Administration.For all other depths, the values sigma and amplitude are linearly interpolated.The source function correction is incorporated in the dose calculation by convolution of the PencilBeam kernels kPB, poly (x, y, d) with the Gaussian distribution gSFC (x, y, d) where the amplitudeA(d) gives the fraction of the Gaussian distribution convoluted with the kernel. Additional measurements are necessary in order to adjust source function correction andradiologic field correction.

Radiologic Field Correction

The radiologic field correction allows small deviations of the radiologic field resulting from gapsettings and the MLC design (round leaf-end and tongue-and-groove) to be corrected with respectto the nominal field size defined by the MLC.• In the leaf direction, an offset can be defined in Physics Administration using the value Leaf

Shift Static in the section Radiologic Field.• Perpendicular to the leaf direction, the required offset can be defined in Physics

Administration using the value Tongue and Groove Size in the section Radiologic Field.

Equivalent Field Size

The MLC equivalent square field size is equal to the square root of the MLC field area, taking intoaccount the radiological field corrections (see the previous section as well as page 63).The jaw equivalent square field size is calculated using the area-to-perimeter formula (refer toSterling et al., 1964 in the general references section of the bibliography).

Idealized Dose Distribution (IDD)

The idealized dose distribution (IDD) for a collimator with an arbitrary shape is the twodimensional convolution of the polyenergetic Pencil Beam kernel with the photon fluence. Itdescribes the dose distribution in photon beams in a homogeneous water phantom and is givenby

IDD x y d, ,( ) φ x' y' d, ,( ) p x' x– y' y– d, ,( ) x′d y′d⋅ ⋅ ⋅=

The photon fluence in an isocenter plane perpendicular to the central beam at a depth d is givenby

φ x y d, ,( ) φ0 x y,( ) RFS r d,( )⋅=

where φ0(x,y) is the fluence matrix in the isocenter plane due to the collimator shape having avalue of 1 for open and 0 for closed fields. Fractional values are used if a matrix voxel is partlycovered by some leaves (see the figure below). RFS(r,d) is the radial factor giving the photonfluence at the following distance from the central beam at a depth d in the phantom:

r x2 y2+=



Fluence Matrix

Figure 5

The fluence matrix (above) shows a contour of the target volume from the beam's eye view.Fractional values are used if matrix voxels are partly covered by leaves.

Total Dose

For calculation of the total dose of a shaped beam at a point in the tissue, the following formulaapplies

D x y d, ,( ) MU NLOut St cmlc c, jaw( )

TPR lrad cd coll,,( )SSDcal dcal+

SSD d+--------------------------------

2

IDD xSID ySID lrad, ,( )

⋅ ⋅ ⋅

⋅ ⋅

=

where:

MU Monitor units applied by the linac.

NLOutNominal linac output, giving the ratio between absolute dose, measured ina water phantom for an open field (calibration field size) at a calibrationdepth dcal, divided by the amount of monitor units (MU) applied.

cjawSize of the jaw equivalent square field, calculated using the area-to-pe-rimeter formula (refer to Sterling et al., 1964 in the general referencessection of the bibilography).

cmlcSize of the MLC equivalent square field, calculated as the square root ofthe MLC field area, taking into account the radiological field corrections.

lradThe radiological path length (depth) of the beam from the tissue surfaceto the observation point, corrected for tissue density inhomogeneities.

SSD Source-surface distance of the central beam.

SID Source-isocenter distance.

d Depth of observation point in tissue.

dcal Depth of the point where NLOut and scatter factors were measured.



St(cmlc,cjaw) Total scatter factor, describing the relative output factor for a squaredMLC and jaw field.

TPR (lrad,cd,coll) Tissue Phantom Ratio, defined as the dose at a point in the phantom div-ided by the dose at the same point at a fixed calibration depth, dcal.

cd c SSD d+SSD

--------------------⋅

IDD(xSID, ySID, Irad)xSID

Idealized dose distribution in depth Irad with

x SIDSSD d+--------------------⋅

and y analogue.

Definition of Coordinates and Parameters

XSID, YSID

lrad

SID

Air

dcal

Pcal

SSD

Source

Tissue

d

(x, y, d)

Figure 6

The convolution between the Pencil Beam kernels and photon fluence map assumes that thePencil Beam kernels are translation invariant in the x and y direction, which means that ahomogeneous medium is assumed. For doses that are calculated next to inhomogeneities, thisassumption may not hold, and the calculation can be incorrect.For inhomogeneous regions traversed by the beam, the correct path length is calculated and thealgorithm computes correct values if the distance to the heterogeneity is large enough thatequilibrium is reestablished.

Radial Factors (RFS)

Radial factors are relative dose distributions along the radial direction of the central beam axis

RFS r d SSD, ,( ) D r d SSD, ,( )D 0 d SSD, ,( )-------------------------------=

and the radial symmetric dose

D r d SSD, ,( ) D x y d SSD, , ,( )=where the radial distance to the central beam is:



r x2 y2+=The other parameters have been described above.

Monitor Unit Calibration

Monitor units (MU) are the unit of measurement used to quantify the dose delivered by a linearaccelerator. These units are calibrated to absorbed dose to water in Gray. This is usuallyperformed using a water phantom under reference conditions at a standard depth, dcal, a source-surface distance, SSDcal and a standard field size (usually 100 x 100 mm²), giving the nominallinac output:

dcal cal MU⁄SSD, ,(cmlc = ccal , cjaw = ccal )NLOut = D

Total Scatter Factor (St)

Total scatter factors (St) describe the relative dose output of a linac at the calibration point in waterfor different jaw and MLC sizes. St account for head and phantom scatter. It is important that thetotal scatter factors St are measured with the calibration setup defined above (SSDcal, dcal).The measurement of total scatter factors is made on the central axis of the beam at the depth dcalin the phantom for various combinations of different square field jaw and MLC sizes.Recommended sizes depend on the type of MLC for which the measurements are made. For thescatter factors doses must be measured at the same depth dcal and the same SSDcal where thenominal linac output is measured. Normalization of the scatter factor St is performed with respectto the dose measured for a certain normalization field size (in general a jaw and MLC field size of100 x 100 mm² is used).

St cmlc cjaw,( )D cmlc cjaw dcal SSDcal, , ,( )D ccal ccal dcal SSDcal, , ,( )---------------------------------------------------------------=

where:

cmlc Size of the square field MLC opening.

cjaw Size of the square field jaw opening.

ccal Calibration field size.

dcalDepth in the phantom where calibration measurements of scatter factors and nomi-nal linac output are performed.

SSDcal Source-surface distance for calibration.

Tissue Phantom Ratio (TPR)

The TPR builds another approach to characterize the depth characteristic of radiation interactions.Compared with the percentage depth dose (PDD), the TPR reflects the more practical situationthat the SSD is changing while the source-detector-distance (SDD) remains constant. It has beenshown [Khan] that the TPR is practically independent of the SSD, since it can be assumed thatthe fractional scatter contribution to the depth dose at a measuring point is only a function of thefield size at the measuring point and the depth of the measuring point in tissue.For calibration of the Pencil Beam algorithm, TPRs are measured by varying the SSD using awater phantom for different (if applicable square) field sizes and a fixed source-detector-distance(SDD).A TPR can also be calculated from a percentage depth dose (PDD) distribution that is measuredwith a constant SSD and the detector moving along the central beam axis. The necessarytransformation is based on the following equation [Khan].



TPR d c, d( )PDD d c SSDcal, ,( )

100-----------------------------------------------

SSDcal d+( )2

SSDcal dcal+( )2----------------------------------------

Sp cdcal( )

Sp cd( )--------------------⋅ ⋅=

using the collimator field size c at the SSD, the collimator field size cd at the depth d

cd cSSDcal d+( )

SSDcal-------------------------------⋅=

and the collimator field size cdcal at the calibration depth dcal

cdcalc

SSDcal dcal+( )SSDcal

-------------------------------------⋅=

Assuming that

Sp cdcal( ) Sp cd( )=≈

the transformation reduces to

TPR d c, d( )PDD d c SSDcal, ,( )

100-----------------------------------------------

SSDcal d+( )2

SSDcal dcal+( )2----------------------------------------⋅=

The error produced through this assumption increases with increasing depth and decreasing fieldsize.NOTE: The tissue phantom ratio is equivalent to the tissue maximum ratio if the calibration depthis equal to the depth of maximum (dcal= dmax).

Radiological Path Length (RPL) Correction

By default, the path length correction is activated for the Pencil Beam algorithm. The algorithmuses ray tracing along a ray ranging from the source to the observation point to calculate theradiological path length. It corrects for tissue inhomogeneities and is based upon CT HounsfieldUnits. It therefore relies on a correct calibration of the CT scanner used for the patient imaging. NOTE: All regions outside the outer contour are assumed to be air and no depth calculation isperformed, independently of the tissue heterogeneity correction setting.

The conversion of CT numbers (HU) to electron density is assumed to be linear in the range from–1000 (electron density = 0.0) up to 47 (electron density = 1.0). Above this value, it is againassumed to be linear but with a different slope. With reference to Schneider 1996 the followingdefault relationship is used:

ρe HU 1000+( ) 1000⁄ 1000– HU 47≤ ≤=

ρe HU 1827.15( ) 1.0213⁄ HU 47>= +

If required, this default can be edited as appropriate using the Physics Administration software.For RPL calculations, the path of a straight ray from the source to a given point inside the patientscan is traced. The distance through every voxel on the ray’s path is scaled by the electron



density of that voxel. The summation of all corrected distances through the voxels gives theradiological path length to the point for which the dose is calculated.

Adaptive Grid Dose Calculation

For calculation of dose distributions in two-dimensional images (i.e. slice views) or in three-dimensional volumes (i.e. DVH) an adaptive grid algorithm is used. The algorithm can acceleratedose calculation tremendously by using the fact that the pixel or voxel resolution is finer than theresolution of the local changes of the dose distribution. The grid size is locally adapted in a way toachieve a predefined accuracy of the dose distribution.The adaptive grid algorithm first calculates the dose values on a coarse grid, using the PencilBeam dose algorithm. Where the dose values in the vicinity of an adaptive grid point can beapproximately described by interpolation, the intermediate dose values between the adaptive gridpoints are interpolated. In the other case, the step size of the adaptive grid points is locallyreduced. The dose values are calculated directly on the new grid points using the dose algorithm.This procedure is repeated recursively until the required accuracy is achieved.As a result, the adaptive grid is typically coarse in regions with smooth dose distributions and finein regions where dose distributions are inhomogeneous (i.e. close to the penumbra region of abeam or close to tissue inhomogeneities).



3.1.1 Pencil Beam for Dynamic Conformal Arc

Dose Calculation

Dynamic arcs are represented as a number of N control points between the start and stop angle.Every control point has its own MLC and jaw shape. For each of the N-1 arc segments, the dosecalculation is discretized by creating one or more segment beams. The segment beams aredistributed uniformly in an arc segment (e.g. in the middle of two control points if one segmentbeam is created for each arc segment).If a single segment beam is created per arc segment, a fluence calculated from the continous leafmovement between the two control points before and after the arc segment is used. If more thanone segment beam is created per arc segment, the MLC and jaw positions are linearlyinterpolated between the two control points before and after the arc segment.The dose delivered by an arc to an arbitrary point is given by the sum of all the arc segmentbeams.

WarningThe calculation assumes that gantry speed and Pencil Beam dose calculation for arctreatments is performed on a discrete gantry angle grid using a finite arc step size (indegrees). Therefore, the calculated dose may be inaccurate and it is highly recommendedto perform a phantom verification for every arc treatment plan.



3.2 Limitations of the Pencil Beam Algorithm3.2.1 Extrapolation Outside the Range of Measured Values

Background

The Pencil Beam dose algorithm uses tabulated measured values for the dose calculation.Although it is not recommended to use these algorithms outside the range of measured values,the extrapolations used by the algorithm are described in the following table. You have to beaware that extrapolated values do not represent reality with the same accuracy as the dosealgorithm generally does.

WarningIf the dose algorithm is used with parameters outside the measured and tabulated values,the accuracy of the calculated dose cannot be guaranteed. You must ensure that allnecessary parameters, in particular the field size, depth and off-axis distance for thepatient treatment are included in the measured beam data.

Measured Values

Depth Dose

Depth < Minimum Depth Constant extrapolation of PDD/TPR (min. depth)

Depth > Maximum Depth

Exponential extrapolation points to determine the exponentialfunction: maximum depth, intermediate depth (depth approxi-mately in the middle between the depth of the maximumdose and maximum depth)

Field Size < Minimum Field Size Constant extrapolation of PDD/TPR (min. field size)

Field Size > Maximum Field Size Constant extrapolation of PDD/TPR (max. field size)

Scatter

MLC Size < Minimum MLC Size Logarithmic extrapolation of scatter

MLC Size > Maximum MLC Size Constant extrapolation of scatter (max. MLC size)

Jaw Size < Minimum Jaw Size Logarithmic extrapolation of scatter

Jaw Size > Maximum Jaw Size Constant extrapolation of scatter (max. jaw size)

RFS

Depth < Minimum Depth Constant extrapolation of RFS (min. depth)

Depth > Maximum Depth Constant extrapolation of RFS (max. depth)

Radius < Minimum Radius Constant extrapolation of RFS (min. radius)

Radius > Maximum Radius Constant extrapolation of RFS (max. radius)

Limitations of the Pencil Beam Algorithm


3.2.2 Other Limitations

Pencil Beam Limitations

The Pencil Beam dose algorithm does not distinguish between the MLC penumbra and the jawpenumbra. Therefore the dose fall-off in the y-direction may be slightly inaccurate for Elekta MLCswith guard leaf behavior (e.g., Agility).

WarningWhen using the Pencil Beam algorithm in dose calculations near inhomogeneous areassuch as lung or bone tissue or close to the tissue border (both within a range of a fewcentimeters), the calculated dose can deviate from the real dose delivered by more than10%.

WarningDepending on the MLC type, the Pencil Beam algorithm uses kernels of a certain resolutionthat define the overall resolution of the dose calculation perpendicular to the beam axis. Inthe case of small structures in combination with a insufficient kernel grid size, the PencilBeam dose calculation may be too coarse to identify every detail of the delivered dosedistribution.

WarningGeneral dose calculation limitations for small treatment fields are summarized in page 151.Ignoring these limitations may lead to deviations of the calculated dose to the real dosedelivered by more than 10%.

Limitations for Small Field Sizes

The Pencil Beam algorithm may be also limited for very small fields due to the influence of thesize and the shape of the electron spot on the bremsstrahlung target. Therefore, for fields smallerthan 10 mm equivalent square field size, it is recommended to use Monte Carlo instead of PencilBeam for dose calculation.



Other Limitations


4 PENCIL BEAM: GENERALBEAM DATAMEASUREMENT

4.1 Introduction

Purpose of This Chapter

This chapter describes the measurement techniques recommended for acquiring the beam datarequired for dose calculation using Brainlab's Pencil Beam algorithm.As well as providing general instructions, this chapter also includes specific information such asMLC and jaw field sizes to be used for the measurements.

Commissioning a Linear Accelerator

Before starting the commissioning of your linear accelerator, you should be familiar with nationalor international recommendations on commissioning a linear accelerator (e.g., the AAPM TG-106Report).This report provides guidelines and recommendations on the proper selection of phantoms anddetectors, setting up a phantom for data acquisition of both scanning and non-scanning data,procedures for acquiring specific photon and electron beam parameters and methods to reducemeasurement errors (< 1%), beam data processing and detector size convolution for accurateprofiles. The procedures described in this report should assist a qualified medical physicist ineither measuring a complete set of beam data, or in verifying a subset of data before initial use orfor periodic quality assurance measurements (Das et al 2008).

Definitions and Abbreviations

Term Explanation

MLC Multileaf Collimator

NLOut Nominal Linac Output

PDD Percentage Depth Dose

RFS Radial Factors

SFC Source Function Correction

SID Source-Isocenter Distance (1000 mm)

SSD Source-surface Distance

TPR

Tissue Phantom RatioNOTE: Depending on the calibration depth dcal, the depth dosedata may actually be TMR (Tissue Maximum Ratio; dcal = dmax).

PENCIL BEAM: GENERAL BEAM DATA MEASUREMENT


Measurement Accuracy

The measurements specified within this user guide are sufficient to achieve the specified accuracyfor Brainlab dose algorithms. If you wish to improve the accuracy of the dose calculation, performthe measurements with extreme care, repeat them, select the best results (e.g., lowest noise) andaverage them. A finer than recommended increment for field size, depth or radial direction,although not prohibited, will not significantly improve dose accuracy.For accurate results, you must set up the linac and the motorized water tank with extreme care.The central beam axis must be exactly vertical, i.e. orthogonal to the water surface. The detectormovement direction must be exactly aligned with the water surface and with the central beam axisin each case.Bear in mind that the sensitivity of the detector may depend on its orientation. Observe thespecifications and recommendations provided by the manufacturer of your dosimetry equipment.Due to the high gradient of flattening-filter free (unflat) beams, it is not recommended to useionization chambers with a cavity volume larger than 0.125 cm3 (e.g. Farmer chambers with avolume of 0.6 cm3) for dose measurements.

WarningThe accuracy of all Brainlab dose algorithms is directly dependent on the accuracy and therange of the beam data measurements. It must be ensured that the beam datameasurement covers the range of field sizes and depths that will be used in subsequenttreatment planning. This is especially the case for the measurements of the scatter factors,the radial profiles and the depth dose.

Data Correction

A limited level of data correction is allowed in order to eliminate small errors during measurementdata acquisition. However, such corrections must be approached with caution. It is always betterto avoid corrections by measuring data that does not need to be modified.

Beam Profile Verification

It is the responsibility of the hospital physicist to perform proper verification of every newly-createdor modified beam profile (machine profile). This must include end-to-end testing for everytreatment modality and treatment condition to be used clinically. You always should consultrelevant national or international recommendations on QA (e.g. IAEA TRS-430).

Responsibility

When measured data is sent to Brainlab, Brainlab has no possibility to verify the correctness of:• any data received from a user• any data returned to a user

Any feedback or recommendations provided by Brainlab based on the data received depend onthe correctness of the data itself. If received data has been processed by Brainlab and returned tothe user, this in no way ensures that the returned data is correct. The user is fully responsible forverifying the correctness of the data returned by Brainlab and is also fully responsible for verifyingthe correctness of any feedback or recommendations provided by Brainlab. The user mustvalidate the safety and effectiveness of the data returned by Brainlab before performing anypatient treatment. The fact that Brainlab may have processed certain data does not affect theoverall responsibility of the user for the correctness of the final beam profile.

Introduction


4.1.1 Recommended Equipment

Background

The following equipment is necessary in order to perform the recommended measurements.Some items are optional and depend on the type of dose algorithm, linac, collimator and treatmentmodality.

Equipment

Component Explanation

Motorized watertank

The tank should extend at least 50 mm beyond all four sides of the measuredfield size at the depth of measurement. It should also extend to at least50 mm beyond the maximum depth of measurement. For a standard field sizeof 400 x 400 mm², a depth of up to 350 mm, a phantom with a base area ofmore than 500 x 500 mm², and a water depth of at least 400 mm is necessa-ry.

Calibratedchamber

A calibrated cylindrical ionization chamber with a cavity volume of at least0.125 cm3 but not more than 0.6 cm3 is required. The effective point of meas-urement shall be determined based on valid international dosimetry standards(e.g. IAEA TRS-398) and the corresponding recommendation of the detectorprovider.

Ionization cham-ber

A cylindrical ionization chamber with a cavity volume of 0.125 cm3 or smaller.The effective point of measurement shall be determined based on valid inter-national dosimetry standards (e.g. IAEA TRS-398) and the corresponding rec-ommendation of the detector provider.

High-resolutiondetector

A very small detector for high-resolution profile measurements and dosimetryof small fields is required. Brainlab recommends the use of an unshieldedstereotactic diode or a single crystal diamond detector.

Plastic phantomThis should be comprised of a number of plastic plates with an electron densi-ty equivalent to the electron density of water. Plastic phantoms should not beused for reference dosimetry.

Radiographicfilm This includes Kodak XV-2 or Kodak EDR-2, and film processing equipment.

Calibrated filmscanner

A calibrated film scanner for film dosimetry and QA (e.g. VIDAR’s Dosimetry-PRO) is required.



4.1.2 General Measurement Requirements

Background

General measurement requirements for all MLCs and linac energies are described below. Specificinformation for particular MLCs is provided on page 65.

Minimum Measurement Requirements

• Nominal Linac Output (NLOut) for a field of 100 x 100 mm² or maximum field size possible.• Leakage for open and closed jaws• Depth dose profiles (TPR/PDD) for different field sizes• Scatter factors (output factors) for different combinations of jaw and MLC sizes• Radial profiles in the diagonal direction for open fields at several depths• Transversal profile measurements for one setup field for adjustment of source function

correction and of radiological field correction.

Additional Measurements

For installations with VMAT functionality, additional dynamic leaf shift measurements may berequired.

Data Input

It is the responsibility of the hospital’s physicist to enter the following data into the PhysicsAdministration:• Nominal Linac output in Gy/100 MU (see page 47)• Source-surface distance (SSD), normalization depth and normalization field size• Leakage values in percent• Parameters for source function correction and radiological field correction (as provided by

Brainlab following processing of customer data)• Depth dose profile table• Scatter factors table• Radial factors table (as provided by Brainlab following processing of customer data)

Further information, refer to the Physics Administration Software User Guide.Optionally, all data can also be entered into the corresponding Excel template available fromBrainlab. The completed template should be sent to [email protected]: It is not possible to mix both methods. For a certain MLC and energy, all data need toeither be collected using the Raw Data method or the Excel method.

General Measurement Requirements



4.2 Absolute Linac Calibration

Nominal Linac Output Measurement

The dose algorithms require as input the relation between monitor units and the absorbed dose towater under reference conditions for a certain beam quality Q (see page 153).This relation is defined as the nominal linac output:

ccal dcal SSDcal,( , ) MU⁄NLOut = D

and is given in Gy/100 MU with:• The calibration field size ccal• The calibration depth dcal• The source-surface distance used for calibration SSDcal

Using the nomenclature of IAEA TRS-398 (V.12), the definition of the nominal linac output is:

w Q, MU⁄=NLOut D

with the reference conditions as defined in table 6.II of IAEA TRS-398 (V.12). The absorbed doseto water Dw,Q is calculated from the dosimeter reading MQ following equation 6.1 of IAEATRS-398 (V.12).

Water Phantom and Chamber Setup

Figure 7

• To measure the nominal linac output you must use the relevant calibrated chamber (see page45).

• The water surface is adjusted at isocenter depth (SSD = SID), with the effective point of theactive chamber volume set at the isocenter (setup position, depth = 0 mm).

• For measurement purposes, the chamber is moved vertically downwards to a depth of 100 mm.



Machine-specific Reference Conditions

Depending on your hardware, it may be possible that reference conditions are not achievable. Ifyour MLC/Jaw combination is not capable of delimiting a reference field fref of 100 x 100 mm2, usethe field size closest to this reference condition as a machine-specific reference field, fmsr (e.g. m3on a Varian linac: typically the jaws are restricted to a 98 x 98 mm2 field, which then defines yourfmsr).The absorbed dose to water for this machine-specific reference field can be calculated followingthe formalism in Alfonso et al 2008, equation 1. If the machine-specific reference field fmsr is veryclose to the reference field fref, then the correction factor

kfmsr fref,

Qmsr Q,

defined in equation 2 can be assumed to be unity and equation 1 of Alfonso et al 2008 can beapproximated by equation 6.1 in IAEA TRS-398 (V.12).

Entering Measurement Results

Enter the source-surface distance (SSD), measurement depth, normalization field size and theNominal Linac Output into Physics Administration in Raw Data mode or, optionally, transferthese data to the Excel template provided by Brainlab and forward the completed template [email protected] you have received the processing results from Brainlab, you can enter the data directly intoPhysics Administration in Machine Profile mode.NOTE: SSD and measurement depth must be identical to the corresponding parameters of thescatter data.

Ensuring Accuracy

To obtain an accurate absorbed dose to water measurement, you must apply a number ofcorrections to the dosimeter reading, e.g. beam quality (linac energy), pressure, temperature andpolarity. You must consult the documentation provided with your dosimetric equipment and thenational standards applicable in your country.

Workflow

Brainlab recommends following a recognized code of practice (e.g. IAEA TRS-398 or AAPMTG-51). Alternatively, the procedure described below may also be used.

Step

1. Set up the motorized water tank with the isocenter at water surface level(SSD = SID = 1000 mm).

2. Adjust the effective point of the active chamber volume to coincide with the isocenter (lev-el of water surface) and mark this as depth zero (see page 47).

3. Move the chamber to the calibration depth of dcal= 100 mm.

4. Set the MLC square field size and the jaw aperture to 100 x 100 mm².NOTE: If your MLC does not facilitate a 100 x 100 mm² field, refer to the instructions spe-cific to your MLC (see page 65).

5. Deliver 100 MU and obtain the dosimeter reading in Gy (apply all necessary conversionsand corrections; e.g. chamber type, beam quality, temperature, pressure, etc.).The result must be in Gy/100 MU.

Absolute Linac Calibration



Step

6. We recommend repeating the measurement three times and using the average value toincrease the accuracy.

NOTE: SSD and normalization depth values other than those specified above may be used.However:

• All the measurements for NLOut, leakage, PDD, radial factors (PDD approach) and scatterfactors must be performed with the same SSD.

• All the measurements for NLOut, leakage and scatter factors must be performed at the samedepth.



4.3 Background Leakage

Open and Closed Jaw Measurement

The background leakage values define the average percentage of leakage radiation through theleaves and through the combination of leaves and jaws.

Setup

• The setup for measuring background leakage is identical to the setup for measuring nominallinac output.

• The same calibrated chamber should be used for measuring nominal linac output andbackground leakage. In this case, the nominal linac output can be used as a reference valuefor determining the leakage value.

• Otherwise, the nominal linac output measurement must be repeated with the new calibrationchamber to obtain this reference value.


Enter leakage values for open and closed jaws into Physics Administration in Raw Data modeor, optionally, transfer these data to the Excel template provided by Brainlab and forward thecompleted template to [email protected] you have received the processing results from Brainlab, you can enter the data directly intoPhysics Administration in Machine Profile mode.

Workflow

Step

1. Leave the primary jaws open at 100 x 100 mm2.

2. Close the MLC leaves asymmetrically.• The leaf gap should be at least 50 mm from the isocenter.• The closed leaf pair gap should be at maximum distance to the isocenter, if the linac

does not support jaw usage.

3. Deliver 100 MU and obtain the dosimeter reading.

4. Close the primary jaws asymmetrically.

5. Repeat step 3.

NOTE: If the linac does not support asymmetric closed jaw settings, the chamber can be movedhorizontally in the x and y directions by at least 20 mm. In this case, the reference value must bemeasured at the new location using a square field jaw and MLC aperture of 100 x 100 mm². If thelinac does not support jaw usage, the jaw leakage should be set to zero.

Background Leakage



4.4 Depth Dose Profile

PDD and TPR Measurement

To determine the depth dose profile, use one of the two options below:

Option Explanation

PDD (percentagedepth dose)

Use this option if the water phantom and SSD are fixed. PDD values aremeasured by adjusting the chamber vertically along the beam axis.

TPR (tissue phantomratio)

Use this option if the chamber is fixed at isocenter. TPR values aremeasured by varying the water level (water surface).

In Physics Administration, you must specify whether your data is based on TPR (isocentricsetup) or PDD (fixed SSD setup). Further information is provided in the corresponding Brainlabplanning software user guide.

Setup

① ②

Figure 8

No. Component

① PDD Measurement (fixed SSD setup)

② TPR Measurement (isocentric setup)

• Large field PDDs shall be measured using an ionization chamber of medium or large volume(0.1 cm3 – 0.6 cm3) to avoid the effect of energy response variations (see section III.D.5 andFig. 1 of the AAPM TG-106 Report (Daset al 2008).

• For smaller field sizes below or equal to 30 mm that are not included in range of field sizes ofthe ionization chamber, use the relevant high-resolution detector (see page 45). In this case, atleast one measurement must be performed with both measurement devices, in order tocompare the measurement results achieved using the ionization chamber with those of thehigh-resolution detector.

TPR Workflow

As Brainlab does not suggest a specific workflow for measuring the TPR, you are free to use aworkflow appropriate to your clinical needs.




Enter the depth dose curves into Physics Administration in Raw Data mode or, optionally,transfer these data to the Excel template provided by Brainlab and forward the completedtemplate to [email protected] you have received the processing results from Brainlab, you can enter the data directly intoPhysics Administration in Machine Profile mode.You must also specify whether your data is based on the TPR (isocentric setup) or the PDD (fixedSSD setup) approach.

Ensuring Accuracy

To prevent errors due to bending of the water surface (capillary effects), move the chamber fromthe bottom upwards.Take care to measure up to a depth that is equivalent to or greater than the maximum depthrequired inside the patient’s body:• For cranial treatments, 250 mm may be sufficient.• For extracranial cases a range of up to 400 mm or more must be covered.

Depending on the clearance of your linac, it may not be possible to measure the TPR up to adepth of 300 mm or 350 mm. If such depths are necessary, you must measure the PDD.NOTE: TPR/PDD values can be normalized arbitrarily as this is controlled by the dose algorithm.

NOTE: Make sure the closed leaf gaps are positioned behind the primary jaws or at a positionaway from the open MLC square field.

PDD Workflow

Step

1. Set up the water phantom in the same way as for the measurement of nominal linac out-put (see page 47):• SSD = SID = 1000 mm• The isocenter is at water surface level (depth = 0 mm)• The effective point of measurement is at isocenter

2. Move the jaws and the MLC leaves to form different square fields for each measurement.The required MLC and jaw field sizes are specified on page 65.

3. Use the water phantom software to measure depths from 0 to the desired depth.Use a step size of 1 mm for 0 - 50 mm depth. For greater depths either a step size of1 mm or a step size of 5 mm may be used.

Depth Dose Profile



4.5 Scatter Factors (Output Factors)

Scatter Factor Measurement

Scatter factors provide information relative to the nominal linac output. They can be entered intoPhysics Administration, normalized arbitrarily, as this is controlled by the dose algorithm.

Daisy-Chain Method

Brainlab recommends following the Daisy-chain method by using an intermediate field size (e.g.,30 x 30 mm²) for the transition between the high-resolution detector for small fields and theionization chamber.

Using the Sample Matrix

• The gray fields in the sample matrix tables provided from page 65 must be measured in allcases.

• The white fields represent MLC and jaw combinations that are not recommended for use withBrainlab’s radiotherapy treatment planning software.

• It is therefore not necessary to measure these larger MLC fields. Instead, it is sufficient to enterthe last mandatory value measured, e.g. in the case of a jaw setting of 60 x 60 mm², you canuse the value measured for the 60 x 60 mm² MLC field (0.8710 on page 86).

• For the Elekta Agility MLC, scatter factors must also be measured for certain MLC field sizeslarger than the corresponding jaw field size (see the sample matrix on page 71). This isnecessary to model guard leaves correctly during Pencil Beam dose calculation.


Enter the complete matrix of scatter factors into Physics Administration in Raw Data mode or,optionally, transfer these data to the Excel template provided by Brainlab and forward thecompleted template to [email protected] you have received the processing results from Brainlab, you can enter the data directly intoPhysics Administration in Machine Profile mode.NOTE: No zero values shall remain in the matrix.

Ensuring Accuracy

• To ensure accuracy, refer to page 21.• Measure the fields exactly as indicated.• Make sure the closed leaf gaps are positioned behind the primary jaws or at a position away

from the open MLC square field.• For each jaw field size, the measured MLC field size range must be larger than or equal to the

jaw field size.

Workflow

Step

1. Set up the motorized water tank with the isocenter at water surface level(SSD = SID = 1000 mm).

2. Adjust the effective point of the active chamber volume to coincide with the isocenter (lev-el of water surface) and mark this as depth zero (see page 47).

3. Move the chamber to the calibration depth of dcal = 100 mm.




Step

4. Measure the scatter factors for a matrix of combinations of square MLC fields and squarejaw fields (see page 86).

The required MLC and jaw field sizes are specified on page 65.

Scatter Factors (Output Factors)


4.6 Diagonal Radial Profiles

Radial Factor and Radial Profile Measurement

Radial factors are dose functions that run horizontally through the beam axis at various depths.Their purpose is to correct off-axis variations of the open beam.If it is not possible to detach the MLC from the gantry head, radial factors cannot be measureddirectly. In this case, radial profiles must be measured diagonally to the jaw field in order to includethe dose fall-off due to the restricted field size.

Figure 9

Radial profiles measured diagonally (to reduce boundary effects) contain the required information,however they must be converted to radial factors using Physics Administration in Raw Data modeor, if Excel Templates are optionally used, at Brainlab headquarters. Radial profile measurementscannot be entered directly into Physics Administration in Machine Profile mode.



Setup

Radial factors/profiles can be measured using an approach similar to that used for TPR and PDDmeasurements.The Beam’s Eye View of measurement direction for Radial Profiles:

Figure 10

• The same water phantom setup as used for TPR/PDD measurements can be used here (seepage 51).

• Make sure the detector moves diagonally from corner to corner. For MLCs with a non-quadraticmaximum field size, the angle of detector movement relative to the leaf direction may differfrom 45°.

• Use the ionization chamber to measure the radial profiles. The chamber should be mounted soas to allow maximum spatial resolution along the measuring direction.

NOTE: The Pencil Beam algorithm assumes rotational symmetry of the radial factors. Therefore, itis possible to measure half profiles, but only if the water phantom is not sufficiently large to allowfor a complete diagonal profile. To improve accuracy, it is recommended to average several halfprofiles measured in different directions.


Enter the diagonal radial profiles into Physics Administration in Raw Data mode or, optionally,transfer your measurement results to the Excel template provided by Brainlab and forward thecompleted template to [email protected] you have received the radial factors from Brainlab, enter them directly into PhysicsAdministration in Machine Profile mode. In Physics Administration you must also specify ifthe isocentric setup or the fixed SSD setup is used.

Ensuring Accuracy

• The radius should correspond with the actual distance between the beam axis and thechamber. Radius conversion to the isocenter plane is not necessary.

Diagonal Radial Profiles



• If radial factors have been measured using an approach similar to that used for PDD, the sameSSD must be used for radial factors as used in the PDD measurement.

• To obtain results with low noise, move the chamber slowly to avoid waves, especially for lowerdepth values.

• Radial profiles can be normalized arbitrarily as this is controlled by the dose algorithm.• For flattened beams, the calculated radial factors are based on the radial profile data ranging

from the beam center to its 50% isodose width. For unflattened beams (FFF mode), thecalculated radial factors are based on the radial profile data ranging from the beam center to its25% isodose width. This beam data range must be measured as exactly as possible for depthsup to and including 200 mm. The scan range for the largest depth is not required to extend upto its 50% (25%) isodose width.

• If measurements at depth 350 mm are not possible due to the limited size of the waterphantom, you can use a smaller depth instead, e.g. 300 mm.

• If half profiles are measured, make sure the beam central axis is NOT too close to the phantomwall. If the beam central axis is too close to the phantom wall, the dose on the central axis isunderestimated resulting in increased horns of the diagonal profiles, especially in larger depths.

Workflow

Step

1. Retract the MLC leaves until the maximum permitted field size is reached (may requirephysics mode of the console).

2. Retract the jaws until the maximum permitted field size is reached (usually same size asthe MLC).

3. Measure the radial profiles for the following depths: 5, 14, 25, 50, 100, 200, and 350[mm].Use a radial resolution of at least 5 mm.

4. Measure the entire profile range from one corner of the field to the other.



4.7 Transversal Profiles

Transversal Profile Measurement

Additional measurements are necessary in order to adjust the source function correction and theradiologic field correction. The source function correction is an empirical way of simulating anextended beam source and other effects that smear out the beam edge.

Water Phantom Measurements

Figure 11

Measurement Options

Two measurement options are provided:• Water Phantom and high-resolution detector: See page 60 (preferred option)• Film Dosimetry: See page 61

Transversal Profiles


Setup

Beam’s Eye View of measurement direction for Transversal Profiles:

Figure 12

The transversal profiles are measured along (x-direction ) AND perpendicular (y-direction) to theleaf direction. You must ensure that the profiles are taken directly under the leaves in such a waythat they are not influenced by the interleaf or intraleaf gap. Refer to the diagrams from page 65on for the correct shape of the corresponding MLC type. The figure above is only a schematic.



4.7.1 Measurement Using a Water Phantom and High-Resolution Detector

Background

It is recommended to perform the transversal profile measurements using a water phantom andhigh-resolution detector. The high-resolution detector as specified on page 45 must be used in thiscase.

Setup

Set up the water phantom in the same way as previously (SSD = SID = 1000 mm).• The measurements should be performed at the following depths: dmax, 100 mm, and 200 mm.• Recalibrate the x/y/z coordinates of the water phantom if the detector has been changed or

rotated.• Use the MLC and the jaws to form the shape specified for your MLC (see page 65).• The transversal profiles shall be measured along (x-direction) AND perpendicular (y-direction)

to the leaf direction.


Import the measured profiles into Physics Administration in Raw Data mode or, optionally, tothe Excel template provided by Brainlab and forward the completed template [email protected]. Enter the source function correction and the radiologic leaf shift inPhysics Administration in Machine Profile mode once you have received the processing resultsfrom Brainlab.

Ensuring Accuracy

Check your measured profiles against the example shown in page 58, and make sure that:• The penumbra width is small (approximately 4±1 mm for micro MLCs)• The outside MLC leakage for a depth less than 50 mm is close to 3%

Workflow

To provide sufficient information about the penumbra region and the area blocked by the MLC, theprofiles should cover the entire field size with a resolution of 0.5 mm.

Step

1. Position the detector so that an interleaf gap is avoided, for example for a MLC designusing 2 central leaves, take the x-profile from the middle of one of the two central leaves.

2. Orient the detector so as to allow maximum resolution for the profile measurement (re-member to rotate the detector between the parallel and perpendicular direction).

3. Position the detector at the beam center and make sure that the distance of the intraleafgaps to the scan axis is more than 20 mm; adjust the axis if necessary.

4. Repeat step 2 with the next measurement depth.

Measurement Using a Water Phantom and High-Resolution Detector



4.7.2 Film Dosimetry Measurement

Background

If measurements using a water phantom and high-resolution detector are not an option, filmdosimetry measurements can be performed instead. For accurate film dosimetry, measurementsshould be performed using film and a plastic phantom comprising variable build-up layers andsufficient back scatter material that is 100 mm thick.

Setup

The measurements should be performed at the following depths: dmax, 100 mm, and 200 mm.• Place the back scatter material on the patient couch and align it to the isocenter position using

the positioning lasers.• Make sure that the upper surface of the back scatter material is horizontally aligned at

isocenter level.• Place a film on the lower slab in the isocenter plane, and add the appropriate build-up layer.• Use the MLC and the jaws to form the shape specified for your MLC (see page 65).


Import the measured profiles into Physics Administration in Raw Data mode or, optionally, tothe Excel template provided by Brainlab and forward the completed template [email protected]. Enter the source function correction and the radiologic leaf shift inPhysics Administration in Machine Profile mode once you have received the processing resultsfrom Brainlab.

Ensuring Accuracy

Check your measured profiles against the example shown in page 58, and make sure that:• The penumbra width is small (approximately 4±1 mm for micro MLCs)• The outside MLC leakage for a depth less than 50 mm is close to 3%

During film dosimetry measurement, the accuracy of the sensitometric curve is essential foraccurate adjustment of the source function correction.

Workflow

• Profiles (x and y) must be extracted for each film.• To provide sufficient information about the penumbra region and the area blocked by the MLC,

the profiles should cover the entire field size with a resolution of 0.5 mm.

Step

1. Expose three films using the build-up layers at dmax (depending on the linac energy, e.g.15 mm for 6 MV), 100 mm and 200 mm.• In order to remain within the linear range of the sensitometric curve, irradiate using an

appropriate level of MU.• For example, for Kodak X-Omat film, the dose should not exceed 0.8 Gy. For Kodak

EDR2 film, the dose should not exceed 2.0 Gy.

2. Develop and scan the films.

3. Extract the x-profile parallel to the leaf direction.To avoid an interleaf gap, take the profile from the middle of one of the central leaves.

4. Extract the y-profile through the beam center perpendicular to the leaf direction.Make sure that the distance of the leaf gaps to the scan axis is more than 20 mm; adjustthe axis if necessary.




4.8 Dynamic Leaf Shift Measurements

Dynamic Leaf Shift

The dynamic leaf shift describes an effective leaf shift due to the round leaf end design of mostMLCs.This value is determined by using Varian MLC files or DICOM files provided by Brainlab tomeasure the isocenter doses for sliding gaps with different widths.The measured dose D can approximately be described by the linear function

agapbgapbDD leak +⋅=+=− )2( δ

where:• gap is the nominal gap width (1, 5, …, 100 mm)• Dleak is the measured MLC leakage• δ is the effective dynamic leaf shift per leaf

After determination of a and b by linear regression, δ is calculated by:

ba 2=δ

Setup

• Position the relevant calibrated chamber or the ionization chamber (see page 45) in the waterphantom, so that the chamber axis is perpendicular to the leaf direction.

• Adjust the water surface level so that the detector is below the build-up region (dmax or deeper)where [SSD = 1000 mm - measurement depth]. For 6 MV, the recommended depth is 20 mm.

• Set the jaws to form a square field of 100 x 100 mm².


All data should also be entered in Physics Administration in Raw Data mode or, optionally, intothe corresponding Excel template available from Brainlab. The result is calculated and displayedusing Physics Administration or the Excel template. Refer to the Physics AdministrationSoftware User Guide for details on how to enter and process the dynamic leaf shift data.

Workflow

Step

1. Successively irradiate the dynamic MLC fields specified for your MLC on page 65.

2. Close the MLC and measure the leakage dose using the same setting as above.With an asymmetric gap setting, the leaf gap should be 50 mm off the isocenter.

3. Set the MLC to a square field of 100 x 100 mm² and measure the open field dose usingthe same settings as above.NOTE: Use the same MU and same dose rate as in step 1.

Dynamic Leaf Shift Measurements


4.9 Verification of Radiologic Field Corrections

Background

This section describes how to verify and update the radiologic field corrections (static anddynamic radiologic leaf shift).The reviewed parameters are employed in the customer specific machine profile, which is used inconjunction with the Brainlab RT Elements treatment planning software.

When to Verify

Brainlab recommends verifying the radiologic field corrections routinely, and especially aftermodifications to the MLC, such as:• Mechanical changes to the infrared light barrier, which calibrates the leaf positions during the

MLC initialization• Exchanging the power supply of an add-on MLC• Changes to the MLC leaf position calibrations

Depending on the modifications, further measurements may be necessary to ensure that thedelivery system works as intended.

Difference Between Radiologic Field and Geometric Field

The figure below illustrates the difference between the radiologic field and the nominal MLC leafpositions. The radiologic field differs from the nominal MLC field in both the x- and y-direction. Thefield difference in the y-direction (tongue-and-groove size) depends mainly on the geometricdesign of the leaf and thus is independent of MLC modifications. Modifications to the MLC in thex-direction might lead to slightly different radiologic field corrections.Illustration of the difference between the radiologic and the geometric MLC field:

Figure 13

About Verification

The radiologic field corrections are defined as correction distances at the isocenter plane. This iswhy the described film measurement is performed isocentrically. For accurate film dosimetry,Brainlab recommends using a plastic phantom which is composed of a build-up layer of 25 mmand sufficient back scatter material that is at least 100 mm thick.



Comparing the Results

Compare the new radiologic leaf shift value with the currently used static leaf shift value inPhysics Administration. If the difference between both values is not negligible, adjust the staticleaf shift in Physics Administration.If the static leaf shift parameter is outdated, the dynamic leaf shift parameter could be inaccurate.In this case, Brainlab recommends repeating the dynamic leaf shift measurements, described onpage 62.If there is a significant difference between the old and the new dynamic leaf shift value, adjust thestatic leaf shift in Physics Administration.Then save and approve the updated machine profile.

Preparation

Step

1. Place the back scatter material on the patient couch and align its upper edge to the iso-center position using the positioning lasers.

2. Make sure that the upper surface of the back scatter material is horizontally aligned atisocenter level.

3. Place a film on the lower slab in the isocenter plane, and add the appropriate build-up lay-er.

4. Use the MLC to form a quadratic (or nearly quadratic) field: 60 x 60 mm2 MLC field.

5. Make sure that the jaw field boarder extends the MLC field by at least 10 mm on eachside: 80x80 mm2 jaw field.

Workflow

To provide sufficient information about the penumbra region and the area blocked by the MLC, theprofile should cover the entire field size with a resolution of at least 0.5 mm.

Step

1. Expose the film using the build-up layer of 25 mm.• In order to remain within the linear range of the sensitometric curve, irradiate using an

appropriate level of MU.• For example, for a Kodak X-Omat film, the dose should not exceed 0.8 Gy. For a Ko-

dak EDR2 film it should not exceed 2.0 Gy. If the film is not sufficiently linear in theconsidered dose range, use calibration films to transform the gray patterns to dose val-ues.

2. Develop and scan the film.

3. Extract the x-profile (direction parallel to the leaf movement).To avoid an interleaf gap, take the profile from the middle of one of the central leaves.

4. Measure the 50% isodose width of the profile and determine the static leaf shift (Δs) be-tween the nominal field size in x-direction (snominal) and the measured 50% isodose width(s50%):

Δs 0.5∗ s50% snominal–( ) Δs 0>( ),=

Verification of Radiologic Field Corrections


5 PENCIL BEAM: BEAMDATA CHECKLISTS

5.1 Beam Data for Brainlab m3

Checklist

Measurement See also Done

Linac calibration (NLOut) using the calibrated chamber:• Jaw: 98 x 98 mm²• MLC: 100 x 100 mm²

NOTE: Enter square field size of 98 mm for Pencil Beam NLOut Square Field Size.

Page 47 ☐

Background leakage for open and closed jaws using the calibrated chamber:Enter SSD, measurement depth, leakage values and NLOut using the Raw Data mode ofPhysics Administration or the Excel template.Open field:• Jaw: 98 x 98 mm²• MLC: 100 x 100 mm²

Page 50 ☐

Depth dose profile (PDD/TPR) using the ionization chamber and high-resolution detector:• MLC (jaw) fields [mm²]

6 x 6 (8 x 8),12 x 12 (14 x 14),18 x 18 (20 x 20),30 x 30 (32 x 32),42 x 42 (44 x 44),60 x 60 (60 x 60),80 x 80 (80 x 80),100 x 100 (98 x 98)Following measurement completion, import the TPR or PDD values using the Raw Datamode of Physics Administration or the Excel template.Remember to use the appropriate detector for small and large field sizes.NOTE: Enter an effective square field size of 98 mm for the largest depth dose profile.

Page 51 ☐

PENCIL BEAM: BEAM DATA CHECKLISTS



Scatter factors using the ionization chamber and high-resolution detector:• Jaw fields

8 x 8, 14 x 14, 20 x 20, 44 x 44, 60 x 60, 80 x 80, 98 x 98 [mm²]• MLC fields

6 x 6, 12 x 12, 18 x 18, 24 x 24, 30 x 30, 36 x 36, 42 x 42, 60 x 60, 80 x 80, 100 x 100 [mm²]A sample matrix is provided on page 67. Following measurement completion, import thescatter factors using the Raw Data mode of Physics Administration or the Excel template.Remember to use the appropriate detector for small and large field sizes with the corre-sponding cross-calibration.NOTE: Enter a square field size of 100 mm for the largest MLC field (row) and 98 mm for thelargest jaw field (column).

Page 51/Page 21 ☐

Diagonal radial profiles using the ionization chamber:• MLC field: 100 x 100 mm2

• Jaw field: 98 x 98 mm2

Following measurement completion, import the diagonal radial profiles using the Raw Datamode of Physics Administration or the Excel template.

Page 55 ☐

Transversal profiles using the high-resolution detector:• Jaw field: 98 x 98 mm²• MLC fields (see page 68)

Following measurement completion, import the transversal profiles using the Raw Datamode of Physics Administration or the Excel template.

Page 58 ☐

Dynamic leaf shift using the calibrated detector:• For the irradiation of the dynamic leaf gaps, use the MLC files: “M3_1.d01”, “M3_5.d01”,

… , “M3_100.d01”.• For each field: Deliver 300 MU at a dose rate of 300 MU/min.

Calculate the dynamic leaf shift using the Raw Data mode of Physics Administration or theExcel template.Use an appropriate combination of m MU and m MU/min if dose rate 300 MU/min is notavailable.

Page 62 ☐

Process the measurement data using the Raw Data mode of Physics Administration or,optionally, forward the completed Excel file either directly to Brainlab headquarters ([email protected]) or to your local support engineer.

☐

Prepare the required beam profile using Physics Administration (Machine Profile mode).Refer to the Physics Administration Software User Guide for details on how the MachineProfile can be prepared based on the Raw Data method or the Excel file method.

☐

Beam Data for Brainlab m3




5.1.1 Additional Information

Sample Matrix

NOTE: The data measured for your linac may differ. Do not use this example clinically.

Brainlab m3: Possible Measured Scatter Factors (Examples Only)

Jaw Settings [mm]

MLC FieldSizes [mm²] 8 x 8 14 x 14 20 x 20 44 x 44 60 x 60 80 x 80 98 x 98

6 x 6 0.601 0.605 0.606 0.607 0.608 0.608 0.609

12 x 12 0.605 0.746 0.759 0.765 0.766 0.766 0.768

18 x 18 0.605 0.756 0.796 0.814 0.815 0.815 0.817

24 x 24 0.605 0.756 0.810 0.838 0.840 0.841 0.842

30 x 30 0.605 0.756 0.810 0.858 0.860 0.860 0.862

36 x 36 0.605 0.756 0.810 0.876 0.878 0.880 0.880

42 x 42 0.605 0.756 0.810 0.880 0.894 0.896 0.897

60 x 60 0.605 0.756 0.810 0.888 0.931 0.936 0.937

80 x 80 0.605 0.756 0.810 0.888 0.931 0.969 0.972

100 x 100 0.605 0.756 0.810 0.888 0.931 0.969 1.000

• The gray fields must be measured in all cases.• The white fields represent MLC and jaw combinations that are not recommended for use with

Brainlab’s radiotherapy treatment planning software.• Blank fields are not permitted (see page 53).



5.1.2 Transversal Profile Shape

Transversal Profile Shape

Brainlab m3: MLC field setup for profile measurements (X and Y Direction):

IEC 1217: -40mm 0mm +40mm

Figure 14

Open Leaves:Leaf #: 2-4, 7-20, 24-25

Closed Leaves

IEC 1217 leaf position -40 mm and +40 mm -50 mm



5.2 Beam Data for Elekta Agility

Checklist

Do not use guard leaves (e.g. no extra open leaf pairs under the jaw adjacent to and in the samephysical position as the last in-field leaf) for the measurement.


Linac calibration (NLOut) using the calibrated chamber.• Y jaw opening: 100 mm• MLC and jaw field size: 100 x 100 mm², no guard leaves

Page 47 ☐

Background leakage for open and closed Y jaw using the calibrated chamber:Enter SSD, measurement depth, leakage values and NLOut using the Raw Data mode ofPhysics Administration or the Excel template.NOTE: Use small off-center leaf openings to enforce jaw openings with closed MLC if jawscannot be positioned independently from the MLC field.

Page 50 ☐

Depth dose profile (PDD/TPR) using the ionization chamber and high-resolution detector:• MLC (jaw) field sizes (no guard leaves) [mm²]([mm]):

10 x 10 (10),20 x 20 (20),30 x 30 (30),40 x 40 (40),60 x 60 (60),80 x 80 (80),100 x 100 (100),140 x 140 (140),200 x 200 (200),300 x 300 (300),400 x 400 (400)Following measurement completion, import the TPR or PDD values using the Raw Datamode of Physics Administration or the Excel template.Remember to use the appropriate detector for small and large field sizes.

Page 51/Page 21 ☐

Scatter factors using the ionization chamber and high-resolution detector:• Y jaw openings

10, 20, 30, 40, 60, 80, 100, 140, 200, 300, 400 [mm]Use small off-center leaf openings to enforce jaw openings larger than the MLC field if jawscannot be positioned independently from the MLC field.• MLC fields, no guard leaves

10 x 10, 20 x 20, 30 x 30, 40 x 40, 60 x 60, 80 x 80, 100 x 100, 140 x 140, 200 x 200,300 x 300, 400 x 400 [mm²]A sample matrix is provided on page 71. Following measurement completion, import thescatter factors using the Raw Data mode of Physics Administration or the Excel template.Remember to use the appropriate detector for small and large field sizes with the corre-sponding cross-calibration.

Page 51/Page 21 ☐

Diagonal radial profiles using the ionization chamber:• Y jaw opening: 400 [mm]• MLC leaves retracted, field size 400 x 400 mm²

NOTE: Measure from corner to corner even if the outermost leaves in the field corners donot fully retract.


Page 55 ☐




Transversal profiles using the high-resolution detector:• Y jaw openings: Y1 jaw at -75 mm, Y2 jaw at 55 mm (5 mm margin to open MLC field)• MLC fields (see page 72)


Page 58 ☐

Dynamic leaf shift using the calibrated detector:For the irradiation of the dynamic leaf gaps, use the corresponding DICOM files.For each field: Deliver 300 MU at a dose rate of 300 MU/min.Use an appropriate combination of m MU and m MU/min if dose rate (300 MU/min) is notavailable.Calculate the dynamic leaf shift using the Raw Data mode of Physics Administration or theExcel template.

Page 62 ☐


☐

Prepare the required beam profile using Physics Administration (Machine Profile mode).Refer to the Physics Administration Software User Guide for details on how the MachineProfile can be prepared based on the Raw Data method or the Excel file method.

☐

Guard Leaves

Do not use Guard Leaves for Elekta Agility.As the Pencil Beam algorithm is based on a lookup table, commissioning without guard leaves ismore accurate when using Brainlab treatment planning for the following reasons:• A 100 x 100 mm² field without guard leaves has a 100 x 100 mm² radiation field (neglecting

radiological shifts), and thus an equivalent square field size of 100 mm. This measurement(e.g., scatter, PDD) is entered using field size 100 mm in Physics Administration.

• A 100 x 100 mm² field with guard leaves has an equivalent square field size which is marginallylarger than 100 mm (the guard leaves marginally increase the total fluence as the jaw-leakageis small, but not zero). If entering such results using equivalent size 100 mm in PhysicsAdministration, a very small error is introduced.

• In Brainlab treatment planning, a 100 x 100 mm² field with guard leaves has an equivalentsquare field size which is marginally larger than 100 mm (the guard leaves marginally increasethe total fluence as the jaw-leakage is small, but not zero). Thus, for a field size slightly largerthan 100 mm, scatter/PDD data would be read from the Pencil Beam data.

• In other words, the effect of the guard leaves (even if it is small) would be considered twiceduring PB dose calculation.

Thus, commissioning without guard leaves works better with the Brainlab table-based PencilBeam algorithm and the field definitions given in this guide.If you wanted to model the same condition for beam measurement as in treatment, you wouldneed to modify the field sizes of the beam data (scatter, PDD table and normalization field size) inPhysics Administration. However, this is more confusing and error-prone than the existingrecommendation (e.g. setting up a 100 x 100 mm² field, which needs to be entered as 100 mm).

Beam Data for Elekta Agility





Sample Matrix


Elekta Agility MLC: Possible Measured Scatter Factors (Examples Only)

Y Jaw Field Size Settings [mm]

MLC FieldSizes [mm²] 10 20 30 40 60 80 100 140 200 300 400

10 x 10 0.6791 0.6946 0.6991 0.7027 0.7063 0.7077 0.7090 0.7099 0.7111 0.7123 0.7124

20 x 20 0.7412 0.8007 0.8080 0.8122 0.8178 0.8198 0.8218 0.8250 0.8259 0.8287 0.8292

30 x 30 0.7491 0.8220 0.8414 0.8469 0.8535 0.8580 0.8611 0.8645 0.8669 0.8695 0.8710

40 x 40 0.7491 0.8372 0.8580 0.8736 0.8821 0.8874 0.8900 0.8944 0.8982 0.9019 0.9035

60 x 60 0.7491 0.8372 0.8580 0.9047 0.9255 0.9332 0.9379 0.9444 0.9496 0.9543 0.9558

80 x 80 0.7491 0.8372 0.8580 0.9047 0.9549 0.9678 0.9733 0.9795 0.9845 0.9879 0.9887

100 x 100 0.7491 0.8372 0.8580 0.9047 0.9549 0.9837 1.0000 1.0087 1.0143 1.0193 1.0198

140 x 140 0.7491 0.8372 0.8580 0.9047 0.9549 0.9837 1.0000 1.0496 1.0573 1.0631 1.0647

200 x 200 0.7491 0.8372 0.8580 0.9047 0.9549 0.9837 1.0000 1.0496 1.0998 1.1075 1.1092

300 x 300 0.7491 0.8372 0.8580 0.9047 0.9549 0.9837 1.0000 1.0496 1.0998 1.1491 1.1511

400 x 400 0.7491 0.8372 0.8580 0.9047 0.9549 0.9837 1.0000 1.0496 1.0998 1.1491 1.1707

• No guard leaves.• Use small off-center leaf openings to enforce jaw openings larger than the MLC field if jaws

cannot be positioned independently from the MLC field.• The gray fields must be measured in all cases.• The white fields represent MLC and jaw combinations that are not recommended for use with

Brainlab’s radiotherapy treatment planning software.• Blank fields are not permitted (see page 53).• In contrast to other MLC types, for the Elekta Agility MLC, scatter factors must be measured

also for certain MLC field sizes larger than the corresponding jaw field size (see gray entriesbelow the diagonal in the scatter table). This is necessary to correctly model guard leaves andthe jaw tracking feature during Pencil Beam dose calculation. In treatment plans for the ElektaAgility MLC, the jaws overlap the MLC field.





Elekta Agility MLC: MLC field setup for profile measurements (X and Y Direction):

Figure 15

Open Leaves:Leaf #: 27-34, 37-44, 49-50

Closed Leaves

IEC 1217 leaf position -25 mm and +75 mm -125 mm (with minimum gap)

• No guard leaves• Open leaves #25 and #52 from -125 mm to -120 mm and auto-adjust the jaws if the y-jaws

cannot manually be set to Y1=-75 mm, Y2=55 mm



5.3 Beam Data for MHI MLC 60

Checklist


Linac calibration (NLOut) using the calibrated chamber. Page 47 ☐

Background leakage for closed MLC field using the calibrated chamber:Enter SSD, measurement depth, leakage values and NLOut using the Raw Data mode ofPhysics Administration or into the Excel template.

Page 50 ☐

Depth dose profile (PDD/TPR) using the ionization chamber and high-resolution detector:• MLC fields

10 x 10, 20 x 20, 30 x 30, 40 x 40, 60 x 60, 80 x 80, 100 x 100, 120 x 120, 150 x 150 [mm²]Following measurement completion, import the TPR or PDD values using the Raw Datamode of Physics Administration or the Excel template.Remember to use the appropriate detector for small and large field sizes.

Page 51 ☐

Scatter factors using the ionization chamber and high-resolution detector:• MLC fields

10 x 10, 20 x 20, 30 x 30, 40 x 40, 60 x 60, 80 x 80, 100 x 100, 120 x 120, 150 x 150 [mm²]A sample matrix is provided on page 74. Following measurement completion, import thescatter factors using the Raw Data mode of Physics Administration or the Excel template.Remember to use the appropriate detector for small and large field sizes with the corre-sponding cross-calibration.

Page 51/Page 21 ☐

Diagonal radial profiles using the ionization chamber:• MLC leaves retracted


Page 55 ☐

Transversal profiles using the high-resolution detector:• MLC fields (see page 75)


Page 58 ☐


☐

Prepare the required beam profile using Physics Administration (Machine Profile mode).Refer to the Physics Administration Software User Guide for the details on how the Ma-chine Profile can be prepared based on the Raw Data method or the Excel file method.

☐






Sample Matrix


MHI MLC 60: Possible Measured Scatter Factors (Examples Only)

Jaw Settings for Physics Administration [mm]

MLC Field Sizes [mm²] 150

10 x 10 0.828

20 x 20 0.934

30 x 30 0.953

40 x 40 0.964

60 x 60 0.977

80 x 80 0.988

100 x 100 1.000

120 x 120 1.010

150 x 150 1.022

• To enter the data into Physics Administration, use the jaw field size setting 150 mm.

Additional Information




MHI MLC 60: MLC field setup for profile measurements (X and Y Direction):

Figure 16

Open Leaves:Leaf #: 6-8, 12-19, 22-25

Closed Leaves

IEC 1217 leaf position -50 mm and +50 mm -78 mm and -77.5 mm



5.4 Beam Data for Novalis

Checklist


Linac calibration (NLOut) using the calibrated chamber:• Jaw: 98 x 98 mm²• MLC: 100 x 100 mm²

NOTE: Enter square field size of 98 mm for Pencil Beam NLOut Square Field Size.

Page 47 ☐

Background leakage for open and closed jaws using the calibrated chamber:Enter SSD, measurement depth, leakage values and NLOut using the Raw Data mode ofPhysics Administration or into the Excel template.Open field:• Jaw: 98 x 98 mm²• MLC: 100 x 100 mm²

Page 50 ☐


6 x 6 (8 x 8),12 x 12 (14 x 14),18 x 18 (20 x 20),30 x 30 (32 x 32),42 x 42 (44 x 44),60 x 60 (60 x 60),80 x 80 (80 x 80),100 x 100 (98 x 98)Following measurement completion, import the TPR or PDD values using the Raw Datamode of Physics Administration or the Excel template.Remember to use the appropriate detector for small and large field sizes.NOTE: Enter an effective square field size of 98 mm for the largest depth dose profile.

Page 51 ☐


8 x 8, 14 x 14, 20 x 20, 44 x 44, 60 x 60, 80 x 80, 98 x 98 [mm²]• MLC fields

6 x 6, 12 x 12, 18 x 18, 24 x 24, 30 x 30, 36 x 36, 42 x 42, 60 x 60, 80 x 80, 100 x 100 [mm²]A sample matrix is provided on page 78. Following measurement completion, import thescatter factors using the Raw Data mode of Physics Administration or the Excel template.Remember to use the appropriate detector for small and large field sizes with the corre-sponding cross-calibration.NOTE: Enter a square field size of 100 mm for the largest MLC field (row) and 98 mm for thelargest jaw field (column).

Page 51/Page 21 ☐

Diagonal radial profiles using the ionization chamber:• MLC field: 100 x 100 mm²• Jaw field: 98 x 98 mm²


Page 55 ☐

Transversal profiles using the high-resolution detector:• Jaw field: 98 x 98 [mm²]• MLC fields (see page 79)


Page 58 ☐

Beam Data for Novalis



Dynamic leaf shift using the calibrated detector:• For the irradiation of the dynamic leaf gaps, use the MLC files: “M3_1.d01”, “M3_5.d01”,

… , “M3_100.d01”.• For each field: deliver 320 MU at a dose rate of 320 MU/min.

Calculate the dynamic leaf shift using the Raw Data mode of Physics Administration or us-ing the Excel template.

Page 62 ☐


☐


☐






Sample Matrix


Novalis: Possible Measured Scatter Factors (Examples Only)

Jaw Settings [mm]

MLC FieldSizes [mm²] 8 x 8 14 x14 20 x 20 44 x44 60 x 60 80 x 80 98 x 98

6 x 6 0.601 0.605 0.606 0.607 0.608 0.608 0.609

12 x 12 0.605 0.746 0.759 0.765 0.766 0.766 0.768

18 x 18 0.605 0.756 0.796 0.814 0.815 0.815 0.817

24 x 24 0.605 0.756 0.810 0.838 0.840 0.841 0.842

30 x 30 0.605 0.756 0.810 0.858 0.860 0.860 0.862

36 x 36 0.605 0.756 0.810 0.876 0.878 0.880 0.880

42 x 42 0.605 0.756 0.810 0.880 0.894 0.896 0.897

60 x 60 0.605 0.756 0.810 0.888 0.931 0.936 0.937

80 x 80 0.605 0.756 0.810 0.888 0.931 0.969 0.972

100 x 100 0.605 0.756 0.810 0.888 0.931 0.969 1.000







Novalis: MLC field setup for profile measurements (X and Y Direction):

IEC1217: -40mm 0mm +40mmFigure 17

Open Leaves:Leaf #: 2-4, 7-20, 24-25

Closed Leaves


References

For further examples of the dosimetric characteristics of Novalis, refer to Yin et al, 2002 (see thegeneral references provided on page 155).



5.5 Beam Data for Varian HD120 (SRS FlatteningFilter)

Checklist

NOTE: SRS mode refers to the SRS-mode flattening filter of Novalis TX and Varian Trilogy linacs.This SRS mode uses 6 MV photon beams and a high dose rate of 1000 MU/min in combinationwith a limited maximum field size of 150 x 150 mm². If the linac is in SRS mode, it uses a differentflattening filter. Hence, beam data for Standard and SRS mode are slightly different.



Background leakage for open and closed jaws using the calibrated chamber:Enter SSD, measurement depth, leakage values and NLOut using the Raw Data mode ofPhysics Administration or into the Excel template.

Page 50 ☐


5 x 5 (8 x 8),10 x 10 (12 x 12),20 x 20 (22 x 22),30 x 30 (32 x 32),40 x 40 (42 x 42),60 x 60 (60 x 60),80 x 80 (80 x 80),100 x 100 (100 x 100),120 x 120 (120 x 120),150 x 150 (150 x 150)Following measurement completion, import the TPR or PDD values using the Raw Datamode of Physics Administration or the Excel template.Remember to use the appropriate detector for small and large field sizes.

Page 51 ☐


8 x 8, 12 x 12, 22 x 22, 32 x 32, 42 x 42, 60 x 60, 80 x 80, 100 x 100, 120 x 120, 150 x 150[mm²]• MLC fields

5 x 5, 10 x 10, 20 x 20, 30 x 30, 40 x 40, 60 x 60, 80 x 80, 100 x 100, 120 x 120, 150 x 150[mm²]A sample matrix is provided on page 82. Following measurement completion, import thescatter factors using the Raw Data mode of Physics Administration or the Excel template.Remember to use the appropriate detector for small and large field sizes with the corre-sponding cross-calibration.

Page 51/Page 21 ☐

Diagonal radial profiles using the ionization chamber:• Jaw field: 150 x 150 [mm²]• MLC leaves parked, field size 150x150 mm2


Page 55 ☐



Page 58 ☐

Beam Data for Varian HD120 (SRS Flattening Filter)



Dynamic leaf shift using the calibrated detector:• For the irradiation of the dynamic leaf gaps, use the MLC files: “NTx_1.d01”, “NTx_5.d01”,

… , “NTx_100.d01”, or the corresponding DICOM files “DynLeafShift - Varian HD120- ...dcm”.

• In each case: deliver 1000 MU at a dose rate of 1000 MU/min.Calculate the dynamic leaf shift using the Raw Data mode of Physics Administration or us-ing the Excel template.

Page 62 ☐


☐


☐






Sample Matrix


Varian HD120 (SRS): Possible Measured Scatter Factors (Examples Only)

Jaw Settings [mm]

MLC FieldSizes [mm²] 8 x 8 12 x12 22 x 22 32 x32 42 x 42 60 x 60 80 x 80 100 x 100 120 x 120 150 x 150

5 x 5 0.638 0.682 0.685 0.686 0.687 0.688 0.689 0.692 0.692 0.694

10 x 10 0.647 0.807 0.818 0.819 0.819 0.820 0.826 0.824 0.826 0.828

20 x 20 0.647 0.815 0.876 0.881 0.891 0.892 0.894 0.894 0.895 0.899

30 x 30 0.647 0.815 0.883 0.892 0.913 0.923 0.924 0.927 0.926 0.929

40 x 40 0.647 0.815 0.883 0.907 0.927 0.939 0.946 0.949 0.949 0.954

60 x 60 0.647 0.815 0.883 0.907 0.935 0.959 0.972 0.974 0.977 0.979

80 x 80 0.647 0.815 0.883 0.907 0.935 0.959 0.982 0.989 0.992 0.996

100 x 100 0.647 0.815 0.883 0.907 0.935 0.959 0.982 1.000 1.007 1.009

120 x 120 0.647 0.815 0.883 0.907 0.935 0.959 0.982 1.000 1.018 1.027

150 x 150 0.647 0.815 0.883 0.907 0.935 0.959 0.982 1.000 1.018 1.045







Varian HD120 (SRS): MLC field setup for profile measurements (X and Y Direction):

Figure 18

Open Leaves:Leaf #: 13-14, 23-38, 43-52

Closed Leaves


References

For further examples of the dosimetric characteristics of Varian HD120, refer to Chang et al, 2008(see the general references provided on page 155).



5.6 Beam Data for Varian HD120 (Standard IrradiationMode and Flattening Filter Free Mode)

Checklist




Page 50 ☐


5 x 5 (8 x 8),10 x 10 (12 x 12),20 x 20 (22 x 22),30 x 30 (32 x 32),40 x 40 (42 x 42),60 x 60 (60 x 60),80 x 80 (80 x 80),100 x 100 (100 x 100),140 x 140 (140 x 140),220 x 220 (220 x 220),300 x 220 (300 x 220)Following measurement completion, import the TPR or PDD values using the Raw Datamode of Physics Administration or the Excel template.Remember to use the appropriate detector for small and large field sizes.

Page 51 ☐


8 x 8, 12 x 12, 22 x 22, 32 x 32, 42 x 42, 60 x 60, 80 x 80, 100 x 100, 140 x 140, 220 x 220,300 x 220 [mm²]• MLC fields

5 x 5, 10 x 10, 20 x 20, 30 x 30, 40 x 40, 60 x 60, 80 x 80, 100 x 100, 140 x 140, 220 x 220,300 x 220 [mm²]A sample matrix is provided on page 86. Following measurement completion, import thescatter factors using the Raw Data mode of Physics Administration or the Excel template.Remember to use the appropriate detector for small and large field sizes with the corre-sponding cross-calibration.

Page 51/Page 21 ☐

Diagonal radial profiles using the ionization chamber:• Jaw field: 400 x 400 [mm²]• MLC leaves parked, field size 400 x 400 mm2


Page 55 ☐



Page 58 ☐

Beam Data for Varian HD120 (Standard Irradiation Mode and Flattening Filter Free Mode)



Dynamic leaf shift using the calibrated detector:• For the irradiation of the dynamic leaf gaps, use the MLC files: “NTx_1.d01”, “NTx_5.d01”,

… , “NTx_100.d01”, or the corresponding DICOM files “DynLeafShift - Varian HD120- ...dcm”.

• For each field: deliver 300 MU at a dose rate of 300 MU/min.Use an appropriate combination of m MU and m MU/min if dose rate 300 MU/min is notavailable.Calculate the dynamic leaf shift using the Raw Data mode of Physics Administration or theExcel template.

Page 62 ☐


☐


☐






Sample Matrix


Varian HD120 (Standard Irradiation Mode): Possible Measured Scatter Factors (Examples Only)

Jaw Settings [mm]

MLCFieldSizes[mm²]

8 x 8 12 x12 22 x 22 32 x32 42 x 42 60 x 60 80 x 80 100 x100

140 x140

220 x220

300 x220(254 x254)

5 x 5 0.6356 0.6673 0.6726 0.6729 0.6739 0.6758 0.6768 0.6791 0.6804 0.6870 0.6860

10 x 10 0.6608 0.7649 0.7754 0.7770 0.7787 0.7800 0.7819 0.7832 0.7852 0.7924 0.7931

20 x 20 0.6608 0.7747 0.8337 0.8412 0.8507 0.8546 0.8572 0.8579 0.8612 0.8677 0.8690

30 x 30 0.6608 0.7747 0.8389 0.8583 0.8782 0.8919 0.8959 0.8978 0.9011 0.9077 0.9090

40 x 40 0.6608 0.7747 0.8389 0.8689 0.8939 0.9122 0.9234 0.9273 0.9306 0.9371 0.9384

60 x 60 0.6608 0.7747 0.8389 0.8689 0.8959 0.9371 0.9542 0.9640 0.9725 0.9797 0.9823

80 x 80 0.6608 0.7747 0.8389 0.8689 0.8959 0.9371 0.9718 0.9856 0.9987 1.0079 1.0105

100 x100 0.6608 0.7747 0.8389 0.8689 0.8959 0.9371 0.9718 1.0000 1.0190 1.0295 1.0314

140 x140 0.6608 0.7747 0.8389 0.8689 0.8959 0.9371 0.9718 1.0000 1.0452 1.0655 1.0655

220 x220 0.6608 0.7747 0.8389 0.8689 0.8959 0.9371 0.9718 1.0000 1.0452 1.1054 1.1133

300 x220(254 x254)

0.6608 0.7747 0.8389 0.8689 0.8959 0.9371 0.9718 1.0000 1.0452 1.1054 1.1244


Brainlab’s radiotherapy treatment planning software.• Blank fields are not permitted (see page 53).• For the scatter factors measured for a rectangular field size of 300 x 220 mm², the equivalent

square field size of 254 mm must be used in Physics Administration.





Varian HD120 (Standard Irradiation Mode): MLC field setup for profile measurements (X and YDirection):

Figure 19

Open Leaves:Leaf #: 13-14, 23-38, 43-52

Closed Leaves


References

For further examples of the dosimetric characteristics of Varian HD120, refer to Chang et al, 2008(see the general references provided on page 155).



5.7 Beam Data for Varian 120 (SRS Flattening Filter)

Checklist





Page 50 ☐


10 x 10 (12 x 12),20 x 20 (22 x 22),30 x 30 (32 x 32),40 x 40 (42 x 42),60 x 60 (60 x 60),80 x 80 (80 x 80),100 x 100 (100 x 100),120 x 120 (120 x 120),150 x 150 (150 x 150)Following measurement completion, import the TPR or PDD values using the Raw Datamode of Physics Administration or the Excel template.Remember to use the appropriate detector for small and large field sizes.

Page 51 ☐


12 x 12, 22 x 22, 32 x 32, 42 x 42, 60 x 60, 80 x 80, 100 x 100, 120 x 120, 150 x 150 [mm²]• MLC fields

10 x 10, 20 x 20, 30 x 30, 40 x 40, 60 x 60, 80 x 80, 100 x 100, 120 x 120, 150 x 150 [mm²]A sample matrix is provided on page 90. Following measurement completion, import thescatter factors using the Raw Data mode of Physics Administration or the Excel template.Remember to use the appropriate detector for small and large field sizes with the corre-sponding cross-calibration.

Page 51/Page 21 ☐

Diagonal radial profiles using the ionization chamber:• Jaw field: 150 x 150 [mm²]• MLC leaves parked, field size 150 x 150 mm2


Page 55 ☐



Page 58 ☐

Beam Data for Varian 120 (SRS Flattening Filter)



Dynamic leaf shift using the calibrated detector:• For the irradiation of the dynamic leaf gaps, use the MLC files: “V120_1.d01”,

“V120_5.d01”, … , “V120_100.d01”, or the corresponding DICOM files “DynLeafShift - Var-ian 120 - ...dcm”.

• In each case: deliver 1000 MU at a dose rate of 1000 MU/min.Calculate the dynamic leaf shift using the using the Raw Data mode of Physics Administra-tion or the Excel template.

Page 62 ☐


☐


☐






Sample Matrix


Varian 120 (SRS): Possible Measured Scatter Factors (Examples Only)

Jaw Settings [mm]

MLC Field Sizes[mm²] 12x12 22x22 32x32 42x42 60x60 80x80 100x 100 120x 120 150x 150

10 x 10 0.800 0.818 0.819 0.819 0.820 0.826 0.824 0.826 0.828

20 x 20 0.807 0.869 0.881 0.891 0.892 0.894 0.894 0.895 0.899

30 x 30 0.807 0.876 0.892 0.913 0.923 0.924 0.927 0.926 0.929

40 x 40 0.807 0.876 0.907 0.927 0.939 0.946 0.949 0.949 0.954

60 x 60 0.807 0.876 0.907 0.938 0.959 0.972 0.974 0.977 0.979

80 x 80 0.807 0.876 0.907 0.938 0.959 0.982 0.989 0.992 0.996

100 x 100 0.807 0.876 0.907 0.938 0.959 0.982 1.000 1.007 1.009

120 x 120 0.807 0.876 0.907 0.938 0.959 0.982 1.000 1.018 1.027

150 x 150 0.807 0.876 0.907 0.938 0.959 0.982 1.000 1.018 1.045







Varian 120 (SRS): MLC field setup for profile measurements (X and Y Direction):

Figure 20

Open Leaves:Leaf #: 21-22, 27-34, 37-44

Closed Leaves




5.8 Beam Data for Varian 120 (Standard IrradiationMode and Flattening Filter Free Mode)

Checklist




Page 50 ☐


10 x 10 (12 x 12),20 x 20 (22 x 22),30 x 30 (32 x 32),40 x 40 (42 x 42),60 x 60 (60 x 60),80 x 80 (80 x 80),100 x 100 (100 x 100),140 x 140 (140 x 140),200 x 200 (200 x 200),300 x 300 (300 x 300)Following measurement completion, import the TPR or PDD values using the Raw Datamode of Physics Administration or the Excel template.Remember to use the appropriate detector for small and large field sizes.

Page 51 ☐


12 x 12, 22 x 22, 32 x 32, 42 x 42, 60 x 60, 80 x 80, 100 x 100, 140 x 140, 200 x 200,300 x 300 [mm²]• MLC fields

10 x 10, 20 x 20, 30 x 30, 40 x 40, 60 x 60, 80 x 80, 100 x 100, 140 x 140, 200 x 200,300 x 300 [mm²]A sample matrix is provided on page 94. Following measurement completion, import thescatter factors using the Raw Data mode of Physics Administration or the Excel template.Remember to use the appropriate detector for small and large field sizes with the corre-sponding cross-calibration.

Page 51/Page 21 ☐

Diagonal radial profiles using the ionization chamber:• MLC leaves retracted, field size 400 x 400 mm2

• Jaws retracted, field size 400 x 400 mm2


Page 55 ☐



Page 58 ☐

Beam Data for Varian 120 (Standard Irradiation Mode and Flattening Filter Free Mode)



Dynamic leaf shift using the calibrated detector:• For the irradiation of the dynamic leaf gaps, use the MLC files: “V120_1.d01”,

“V120_5.d01”, … , “V120_100.d01”, or the corresponding DICOM files “DynLeafShift - Var-ian 120 - ...dcm”.

• For each field: deliver 300 MU at a dose rate of 300 MU/min.Use an appropriate combination of m MU and m MU/min if dose rate 300 MU/min is notavailable.Calculate the dynamic leaf shift using the Raw Data mode of Physics Administration or theExcel template.

Page 62 ☐


☐


☐






Sample Matrix


Varian 120 (Standard Irradiation Mode): Possible Measured Scatter Factors (Examples Only)

Jaw Settings [mm]

MLC FieldSizes[mm²]

12 x 12 22 x 22 32 x 32 42 x 42 60 x 60 80 x 80 100 x 100 140 x 140 200 x 200 300 x 300

10 x 10 0.6730 0.6790 0.6800 0.6810 0.6800 0.6810 0.6840 0.6860 0.6920 0.6920

20 x 20 0.6760 0.7820 0.7840 0.7870 0.7870 0.7880 0.7920 0.7930 0.8010 0.8010

30 x 30 0.6760 0.7830 0.8030 0.8230 0.8260 0.8270 0.8310 0.8330 0.8420 0.8420

40 x 40 0.6760 0.7830 0.8090 0.8350 0.8490 0.8530 0.8560 0.8580 0.8660 0.8660

60 x 60 0.6760 0.7830 0.8090 0.8400 0.8710 0.8800 0.8880 0.8930 0.8980 0.8980

80 x 80 0.6760 0.7830 0.8090 0.8400 0.8710 0.8940 0.9080 0.9130 0.9190 0.9190

100 x 100 0.6760 0.7830 0.8090 0.8400 0.8710 0.8940 0.9200 0.9250 0.9320 0.9320

140 x 140 0.6760 0.7830 0.8090 0.8400 0.8710 0.8940 0.9200 0.9410 0.9520 0.9600

200 x 200 0.6760 0.7830 0.8090 0.8400 0.8710 0.8940 0.9200 0.9410 0.9770 0.9820

300 x 300 0.6760 0.7830 0.8090 0.8400 0.8710 0.8940 0.9200 0.9410 0.9770 1.0050







Varian 120 (Standard Irradiation Mode): MLC field setup for profile measurements (X and YDirection)

Figure 21

Open Leaves:Leaf #: 21-22, 27-34, 37-44

Closed Leaves






6 MONTE CARLO:ALGORITHM

6.1 Introduction to the Monte Carlo Algorithm

General Overview

The description that follows shall provide an overview of the physical features behind the BrainlabMonte Carlo (MC) algorithm. The intention is to allow the user to work with the software, tounderstand the behavior of the MC algorithm and to understand the meaning of the MC useroptions. For more detailed information about the MC techniques in general and XVMC inparticular, we refer you to the publications listed on page 155.New cancer treatment techniques such as IGRT or VMAT allow more precise dose deposition inthe target volume and an improved control of the normal tissue complications. An accurate dosecalculation is essential to assure the quality of the improved techniques. Conventional dosecalculation methods, such as the Pencil Beam algorithm, are of high quality in regions withhomogeneous tissue, e.g. within the brain. However, for treatments in the head-and-neck or in thethorax regions, i.e. in regions consisting of bone, soft tissue and air cavities, an improvedaccuracy is required. For example, the Pencil Beam algorithm is known to overestimate the dosein the target volume for the treatment of small lung tumors because the Pencil Beam algorithmcalculates dose by scaling the Pencil Beam dose distribution kernels in water to take the tissueheterogeneities into account. This method has accuracy limitations in these regions. MC dosecalculation algorithms, on the other hand, provide more accurate results, especially inheterogeneous regions.

Usage of Monte Carlo in Radiotherapy

In radiotherapy, MC techniques are applied to solve the transport problem of ionizing radiationwithin the human body. Here the radiation is decomposed into single quantum particles (photons,electrons, positrons). The motion of these particles through the irradiation device and the humantissue is simulated by taking into account the material properties of the different components ofthe linac head and the tissue properties in each volume element (voxel). The photons, electronsand positrons interact with the electrons of the atomic shells and the electromagnetic field of theatomic nuclei. This can cause ionization events. The corresponding interaction properties arebased on the laws of quantum physics. For the linac head, these properties can be calculatedusing the known atomic composition of the different components; for the patient they can becalculated based on the CT images and the Hounsfield Unit in each voxel. The interactionproperties are given as total and differential cross sections. Total cross sections characterize theinteraction probabilities of a particle with a given energy in a medium with a definite atomiccomposition. Differential cross sections characterize the probability distribution functions for thegeneration of secondary particles with definite secondary particle parameters, such as energy andscattering angle. The random numbers in a MC simulation are required to sample the specificparameters from these probability distribution functions.For a more thorough introduction into all issues associated with clinical implementation of MonteCarlo-based external beam treatment planning, we refer to the review by Reynaert et al (2007), orthe AAPM Task Group Report No 105 (2007).

MONTE CARLO: ALGORITHM


6.1.1 Brainlab Monte Carlo Algorithm

Background

The Brainlab Monte Carlo algorithm is based on the X-ray Voxel Monte Carlo algorithm developedby Iwan Kawrakow and Matthias Fippel (Kawrakow et al 1996, Fippel et al 1997, Fippel 1999,Fippel et al 1999, Kawrakow and Fippel 2000, Fippel et al 2003, Fippel 2004).

Brainlab Monte Carlo Algorithm

The three components of the Brainlab MC algorithm:

Figure 22

The Brainlab MC algorithm consists of three main components. The first component is used as theparticle source. It models the upper part of the linac head (target, primary collimator, flatteningfilter) and generates photons as well as contaminant electrons from the corresponding distribution.The particles are then transferred to the second component, the model of the collimating system.Depending on the field configuration, the particles are absorbed, scattered or passed through thecollimator system without interaction. The surviving particles are transferred to the patient dosecomputation engine. In this third component, the radiation transport through the patient geometryis simulated and the dose distribution is computed. In the following sections, the threecomponents of the Brainlab MC algorithm are characterized in further detail.

Brainlab Monte Carlo Algorithm


6.2 The Virtual Energy Fluence Model (VEFM)

Background

The geometry of the target, the flattening filter and the primary collimator do not change when thefield shape is changed. Therefore, it can be assumed that the phase space of photons andcharged particles above the jaws and the multileaf collimators (MLC) is independent of the fieldconfiguration. To model this phase space, a Virtual Energy Fluence Model (VEFM) is employed.With some extensions this model is based on the work by Fippel et al (2003).

Geometry Parameters

The VEFM consists of two or three photon sources with two-dimensional Gaussian shapes andone charged particle (electron) contamination source. The photon sources model Bremsstrahlungphotons created in the target and Compton photons scattered by the primary collimator andflattening filter materials. For the photon sources various parameters are required. For example,the distances of the sources to the nominal beam focus is either estimated or taken from thetechnical information provided by linac vendor. The Gaussian widths (standard deviations) as wellas the relative weights of the photon sources are fitted using measured dose distributions in air.Additional horn correction parameters are also fitted from these measurements. They modeldeviations of the beam profile from an ideal flat profile.

Energy Parameters

The VEFM also requires information about the photon energy spectrum as well as the fluence ofcharged particle contamination at the patient’s surface.This information is derived from a measured depth dose curve Dmeas(z) in water for the referencefield size (field size used for the dose – monitor unit calibration).The curve Dmeas(z) is used to minimize the squared difference to a calculated depth dose curveDcalc(z). Based on the model assumptions, Dcalc(z) is given by:

Dcalc z( ) wγ p E( )Dmono E z,( ) EdEmin

Emax

weDe z( )+=

The set of mono-energetic depth dose curves Dmono(E,z) in water can be calculated using thewhole MC system and the geometric beam model parameters derived after fitting the measuredprofiles in air.

Energy Parameter Range

The set must be calculated for a table of energies reaching from the minimum energy of thespectrum Eminup to an energy that is a little larger than the maximum energy Emax. This allows usto also use Emax as a fitting parameter. In contrast to the original paper (Fippel et al 2003), wemodel the energy spectrum p(E) by:

p E( ) N 1 lE–( )exp–( ) bE–( )exp= Emin E Emax≤ ≤

This function is comparable to spectra calculated using EGSnrc (Kawrakow 2000) and BEAM(Rogers et al 1995), especially in the low energy region.The free parameters l,b and the normalization factor N have to be fitted. For Emin and Emax weusually take fixed values, but it is also possible to adjust them, because sometimes the maximumenergy of the spectrum can be different from the nominal photon energy setting in MV. The



parameter wɣ is the total weight of all photon sources. It is calculated by wɣ= 1 - we with we beingthe weight of the electron contamination source. The parameter we is also fitted using themeasured depth dose in water and the formula on Dcalc(z).It requires the depth dose MC computation of a pure electron contamination source in water De(z).Because most of the electrons originate in the flattening filter, the location of the electron source isassumed to be the foot plane of the filter. The energy spectrum of the electrons is estimated by anexponential distribution as described by Fippel et al (2003).

Monte Carlo and Pencil Beam

During the commissioning procedure all parameters (fixed and fitted) are written into a doseprofile file. This file is then linked to the machine profile of the corresponding linac. It is also linkedto the dose profile for the Pencil Beam algorithm. This means the MC dose calculation algorithmcannot be used without the pencil beam algorithm. This is a restriction, but it has beenimplemented into the Brainlab RT treatment planning software because of three main advantages:• It allows the user to cross check the results using two almost independent dose calculations.• It provides a smooth transition from clinical experience (protocols) based on Pencil Beam dose

calculations to a more accurate experience based on Monte Carlo dose calculations.• The faster Pencil Beam algorithm can be used for the intermediate planning process. Later the

user can switch to Monte Carlo to fine tune the treatment plan.Therefore, commissioning of the MC dose calculation algorithm requires commissioning of thePencil Beam algorithm.Before using the MC dose profile clinically, it has to be validated against measured dosedistributions and output (scatter) factors in a homogeneous water phantom. We refer to the MonteCarlo base data measurement instructions for more information on this data.

The Virtual Energy Fluence Model (VEFM)


6.3 Modeling of the Collimating System

Jaws

The components of the collimating system (jaws and MLC) are modeled in different ways. Therectangle given by the positions of both jaw pairs are used to define the sampling space of theinitial particles. That means only photons and electrons are generated going through the jawopening. In other words, the MC algorithm assumes fully blocking jaws. The error of thisassumption is estimated to be below 0.5% because of the jaw thickness and the attenuation of thejaw material. Furthermore, the beam is also blocked by the MLC, leading to further reduction ofthe photon fluence outside the beam limits. The advantage of this approach is that it savescomputation time. The simulation of photon histories being absorbed within the jaw material wouldjust be a waste of computing power and it would not have a significant effect on the calculationaccuracy.

MLC Leaf Designs

Different MLC leaf designs:

① ② ③

④ ⑤ ⑥

Figure 23

No. Component

① Ideal MLC (no leakage radiation)

② Tilted leaves (Siemens)

③ Step design (Elekta)

④ Tongue and groove design (Varian)

⑤ Varian Millennium

⑥ Brainlab m3

Only 4 leaf pairs per MLC are represented.

MLC Simulation

Depending on the MLC type, the model of the MLC takes into account the correct thickness of theMLC, the widths of the leaves, the material of the leaves, the rounded leaf tips (if available) andthe correct tongue and groove design (see the figure above).

MLC Simulation Algorithm

The algorithm behind these models is based on the work published by Fippel (2004). It is a fullMC geometry simulation of the photon transport. It takes into account Compton interactions, pair



production events and photoelectric absorptions. Primary and secondary electrons are simulatedusing the continuous slowing down approximation. In this approach the geometries are defined byvirtually placing planes and cylinder surfaces in the 3D space. The planes (and surfaces) definethe boundaries between regions of different material. For MLCs, in general the regions consist ofa tungsten alloy and air. For these materials photon cross section tables pre-calculated using thecomputer code XCOM (Berger and Hubbell 1987) as well as electron stopping power and rangetables pre-calculated using the ESTAR software (Berger 1993) are used. The particle ray-tracingalgorithm is based on bit masks and bit patterns to identify the region indices. In extension to theoriginal paper, further MLC models have been implemented.

Determination of Leaf Positions

The Brainlab RT treatment planning software defines the nominal leaf position in the isocenterplane. The real leaf position as required by the MC algorithm is calculated from the nominalposition assuming in most cases a light field calibration. For an MLC with rounded leaf tip thismeans, the leaf touches the straight line between the nominal focus and the nominal leaf position.In this way, the straight line becomes a tangent of the leaf tip curvature. In case of the ElektaAgility MLC on the other hand, the calculation of the real leaf positions is based on a radiologicfield calibration. This requires an additional leaf shift relative to the light field calibration.The leaf positions for closed leaf pairs are determined differently. In the case of closed leaf pairs,the straight line between the nominal focus and the nominal leaf position goes directly through theend point of the leaf tip. This behavior has been implemented to avoid air gaps between oppositeleaf tips if the leaf pair is closed far off axis.

Modeling of the Collimating System


6.4 The MC Patient Dose Computation Engine

Background

The MC algorithm to simulate the transport of photons and electrons through human tissue isbased on the publications by Kawrakow et al (1996), Fippel (1999), Kawrakow and Fippel (2000).It is a condensed history algorithm with continuous boundary crossing to simulate the transport ofsecondary and contaminant electrons. It takes into account and simulates delta electrons (freesecondary electrons created during electron-electron interactions) as well as Bremsstrahlungphotons. For the MC photon transport simulations, Compton interactions, pair production eventsand photoelectric absorptions are considered. Several variance reduction techniques such aselectron history repetition, multiple photon transport or Russian Roulette, speed up the dosecomputation significantly compared to general-purpose MC codes, e.g. EGSnrc (Kawrakow 2000).The MC particle histories can run in parallel threads, therefore the code fully benefits from the useof multi-processor machines. Gantry rotations (static and dynamic) are simulated continuously.This feature is a big advantage compared to other algorithms such as the Pencil Beam, becausethey need discrete gantry positions to model the rotation.

Mass Density Calculation

For MC simulations, knowledge of the photon cross sections as well as the electron collision andradiation stopping powers is required. In general, these parameters can be calculated if the atomiccomposition of the tissue type in each voxel is known. HU numbers are available from the CTimages. However, there is no mapping between the HU and certain tissue types because therelation between the HU and the elemental composition is not unique. In other words, there aredifferent tissue types with the same HU. Therefore, it can cause errors to determine the elementalcomposition in the specific voxel if a HU is only measured using the CT.Instead, Brainlab Monte Carlo uses a different approach based on the CT calibration curve. TheHU to electron density (ED) conversion table converts CT specific HU numbers into tissue specificED numbers. The ED values are relative to water. The MC algorithm internally converts these EDsinto all other parameters required for MC calculations. These are for example, mass densities,photon cross sections and electron stopping and scattering powers. This conversion is performeddirectly, that is, without the additional intermediate step of calculating the elemental composition.The publications about VMC/XVMC referenced on page 155 contain more information about thisapproach.If the mass density ρ is known in a specific voxel, the total cross section for e.g. Comptoninteractions μc(ρ,E) for a photon with energy E can be calculated by:

μC ρ E,( ) ρρW------- fC ρ( ) μC

W E( )=

The function μcw(E) is the tabulated Compton cross section in water, ρw is the mass density ofwater and the function fc(ρ) is a fit function based on analyzing ICRU cross section data for bodytissues (ICRU 1992). The factorization into a function depending only on ρ and a second functiondepending only on E is an approximation. However, the data of ICRU Report 46 (1992) imply thatthis approximation is possible for human tissue.

WarningIn some cases, the HU values of the CT scan do not represent the real characteristics of amaterial (e.g. mass density and material composition). This may lead to inaccurate dosecalculation for non-human tissue materials (e.g. implants).



Compton Cross Section Ratio

Figure 24

The illustration above shows the Compton cross section ratio versus mass density for all materialsof ICRU report 46 (crosses). The line represents a fit to this data. This function is used by MC tocalculate the Compton cross-section.The line in the figure above represents a fit to this data. It is given by:

fc ρ( )0.99 0.01ρ /ρw, ρ ρw≤+

0.85 0.15 ρ ρw⁄ , ρ ρw≥+

≈

This fit function is used by MC to calculate the Compton cross-section. There are a few materialswith deviations between the real cross section ratio and the fit function of up to 1.5%. However,these are materials such as gallstone or urinary stones. Furthermore, the correct elementalcomposition in a given voxel is unknown. Only the HU number is known and different materialcompositions can lead to the same HU. Therefore, the HU number itself has some uncertaintyoverlaying in this manner the uncertainty of the fit function. The influence of the HU numberuncertainty on Monte Carlo calculated dose distributions has been discussed in the literature(Vanderstraeten et al 2007). Similar fit functions exist to calculate the pair production andphotoelectric cross sections as well as the electron collision and radiation stopping powers. Theirdependencies on the mass density of course differ from fc(ρ).The function fc(ρ) is also used to convert mass densities ρ into electron densities ne or vice versa.The relation is given by:

ne neW ρρw------ fC ρ( )=

with new being the electron density of water.

The MC Patient Dose Computation Engine


6.5 MC Parameters

Background

The software user has influence on the MC dose calculation accuracy, the dose calculation timeand the dose result type. This can be done using the MC parameters provided by the Brainlab RTElements. Three parameters are available:• Spatial resolution (in mm)• Statistical uncertainty (in %)• Dose result type (dose to medium or dose to water)

Refer to the Software User Guide of the corresponding Brainlab Element for details on whichoption is adjustable and how it can be changed.

Spatial Resolution

The spatial resolution defines the size of the internal MC dose computation grid. It does not meanhowever that the final MC grid size is exactly equal to the value of the parameter. The MC voxelsare constructed by combining an integer number of pixels from the original CT cube. Therefore,the final sizes of the voxels are only approximately equal to the value of the spatial resolutionparameter. They can also be different for the 3 spatial directions. Furthermore, they cannot besmaller than the initial pixel sizes. The selection of this parameter has a strong influence on thecalculation time. Decreasing this parameter by a factor of 2 can increase the calculation time by afactor of about 6. The spatial resolution is limited to an application-dependent range of values. MCdose calculations for small tumors, should be performed with a spatial resolution of 1-2 mm.

Statistical Uncertainty

The statistical uncertainty parameter estimates the number of particle histories needed to achievethis uncertainty per treatment element (beam or arc) in % of the maximum dose of that treatmentelement. Because everything here is normalized per beam or arc, the final uncertainty in the PTVcan be smaller. For example, if we have 5 overlapping arcs in the PTV and each arc is calculatedwith 2% statistical uncertainty, then the statistical uncertainty in the PTV is about 1%.In the non-overlapping regions, it remains 2%. The statistical uncertainty per voxel decreases withincreasing number of histories Nhist as:

1 Nhist( )

i.e., the statistical uncertainty can be decreased by a factor of 2 if the number of histories isincreased by a factor of 4. Therefore, also the calculation time is increased by a factor of 4. Thedefault setting depends on the application.



Dose Result Type

①

②

Figure 25

The MC algorithm allows the calculation of two different dose types, dose to medium and dose towater. To illustrate the difference, consider one volume element (voxel) of the MC calculation grid(①). The size of one voxel is determined by the CT pixel size and the CT slice thickness as wellas the spatial resolution parameter (see above). For example, one voxel can be as small as 1.0mm³. During a MC simulation, the energy absorbed per voxel is calculated (i.e., the spatialsimulation resolution is given by the voxel size). However, biological structures (e.g., cells) can bemuch smaller than that. In the volume element example above, a cell is represented by the smalloval cavity (②). Using the default setting Dose to medium, the energy absorbed in the voxel isdivided by the mass of the voxel to calculate the energy dose. This energy dose is equivalent tothe energy dose within the cell, if the cell is made of the same tissue as the whole voxel.The situation changes if the tissue type within the cell differs from the average tissue within thevoxel (e.g., a bone marrow cell surrounded by bony tissue). The higher mass density of the bonetissue causes a higher fluence of secondary electrons crossing the soft tissue cell. Consequently,the energy dose within the cell is increased compared to the average dose within the voxel. Thissituation can be modeled using the MC algorithm if you select Dose to water instead of Dose tomedium. Dose to water means energy absorbed in a small cavity of water divided by the mass ofthat cavity, whereas some tissue (e.g., bone) surrounds the cavity.The relation between dose to water, Dw anddose to medium, Dm is calculated by:

with

being the unrestricted electron mass collision stopping the power ratio of water to that of themedium averaged over the spectrum of the photon beam. This ratio is approximately 1.0 for softtissues with a mass density of about 1.0 g/cm³. It increases up to ~1.15 for bony tissues of a massdensity up to 2.0 g/cm³.Therefore, there is no visible difference between dose to medium and dose to water for mosthuman soft tissue types. However, dose to water can be up to 15% larger compared to dose to

MC Parameters


medium for bony tissues (AAPM 2007). Dose to water should be selected if you want to know thedose in small soft tissue cavities embedded within a bony structure. Dose to medium should beselected if you want to know the average dose within the whole voxel.

Monte Carlo Limitations

WarningThe accuracy of the Monte Carlo dose calculation depends on the defined statisticaluncertainty. The value used for final treatment plan approval must be as low as possibleand not higher than 2%. You must keep in mind that the statistical uncertainty is definedrelative to the highest dose per treatment element (beam or arc). As a consequence, theuncertainty of the dose values inside OARs in the low dose region is higher relative to themaximum dose inside the OAR.

WarningGeneral dose calculation limitations for small treatment fields are summarized in page 151.Ignoring these limitations may lead to deviations of the calculated dose to the real dosedelivered by more than 10%.



MC Parameters


7 MONTE CARLO: GENERALBEAM DATAMEASUREMENT

7.1 Introduction

Purpose of This Chapter

This chapter describes the measurement techniques recommended for acquiring the beam datarequired for dose calculation using the Brainlab Monte Carlo algorithm. As well as providinggeneral instructions, this chapter also includes specific information such as MLC and jaw fieldsizes to be used for the measurements.

Commissioning a Linear Accelerator

Before starting the commissioning of your linear accelerator you should be familiar with national orinternational recommendations on commissioning a linear accelerator (e.g., the AAPM TG-106Report):This report provides guidelines and recommendations on the proper selection of phantoms anddetectors, setting up a phantom for data acquisition of both scanning and non-scanning data,procedures for acquiring specific photon and electron beam parameters and methods to reducemeasurement errors (<1%), beam data processing and detector size convolution for accurateprofiles. The procedures described in this report should assist a qualified medical physicist ineither measuring a complete set of beam data, or in verifying a subset of data before initial use orfor periodic quality assurance measurements (Das et al 2008).

Definitions and Abbreviations

Term Explanation

CAX Central Axis

Linac Linear Accelerator

MLC Multileaf Collimator

PDD Percentage Depth Dose

SID Source-Isocenter Distance (1000 mm)

SSD Source-Surface Distance

Prerequisites

The Monte Carlo dose calculation algorithm cannot be used without the Pencil Beam algorithm.Therefore, commissioning of the Monte Carlo dose calculation algorithm requires commissioningof the Pencil Beam algorithm. For further information, see page 65.

MONTE CARLO: GENERAL BEAM DATA MEASUREMENT


Purpose of the Measurements

All measured data in water and air are not used directly during Monte Carlo dose calculation. Themeasurements in air are used intermediately during data processing. They are required mainly tofit deviations from an ideal flat profile (horns, etc.) and to determine the shapes and the weights ofthe different photon sources. Volume effects due to the size of the chamber including the brassbuild-up cap are taken into account during this fitting. The parameters influencing the sizes of thephoton sources and the profile penumbra widths are not fitted using the in-air measurements, theyare adjusted using the small field profile scans for SSD=900 mm in water.The parameters of the energy spectrum and the weight of the electron contamination source arefitted using the SSD=1000 mm, 100x100 mm² depth dose curve in water. The Dose/MUcalibration is also adjusted using the absolute dose for this field. All SSD=900 mm data (includingthe absolute dose) are used for comparison with phantom dose calculations, that is for verificationof the final linac head model. The results of the comparison are plotted into a PDF file. Togetherwith the fitted parameter file (called Monte Carlo Dose Profile), this PDF file is provided for review.

Measurement Accuracy

The measurements specified within this user guide are sufficient to achieve the specified accuracyfor Brainlab dose algorithms. If you wish to improve the accuracy of the dose calculation, performthe measurements with extreme care, repeat them, select the best results (e.g., lowest noise) andaverage them. A finer than recommended increment for field size, depth or radial direction,although not prohibited, will not significantly improve dose accuracy.For accurate results, you must set up the linac and the motorized water tank with extreme care.The central beam axis must be exactly vertical, i.e. orthogonal to the water surface. The detectormovement direction must be exactly aligned with the water surface and with the central beam axisin each case.Bear in mind that the sensitivity of the detector may depend on its orientation. Observe thespecifications and recommendations provided by the manufacturer of your dosimetry equipment.

WarningThe accuracy of all Brainlab dose algorithms is directly dependent on the accuracy and therange of the beam data measurements. It must be ensured that the beam datameasurement covers the range of field sizes and depths that will be used in subsequenttreatment planning. This is especially the case for the measurements of the scatter factors,the radial profiles and the depth dose.

Beam Profile Verification

It is the responsibility of the hospital physicist to perform proper verification of every newly-createdor modified beam profile (machine profile). This must include end-to-end testing for everytreatment modality and treatment condition to be used clinically. You always should consultrelevant national or international recommendations on QA (e.g. IAEA TRS-430).

Responsibility

When measured data is sent to Brainlab, Brainlab has no possibility to verify the correctness of :• any data received from a user• any data returned to a user.

Any feedback or recommendations provided by Brainlab based on the data received depend onthe correctness of the data itself. If received data has been processed by Brainlab and returned tothe user, this in no way ensures that the returned data is correct. The user is fully responsible forverifying the correctness of the data returned by Brainlab and is also fully responsible for verifyingthe correctness of any feedback or recommendations provided by Brainlab. The user mustvalidate the safety and effectiveness of the data returned by Brainlab before performing anypatient treatment. The fact that Brainlab may have processed certain data does not affect theoverall responsibility of the user for the correctness of the final beam profile.

Introduction


7.1.1 Recommended Equipment

Equipment

Component Explanation

Motorized watertank

Use a motorized water tank of sufficient size: the tank should extend at least50 mm beyond all four sides of the measured field size at the depth of meas-urement. It should also extend to at least 50 mm beyond the maximum depthof measurement. For a standard field size of 400 x 400 mm2 up to a depth of350 mm, a phantom of more than 500 x 500 mm2 base area and a waterdepth of at least 400 mm is necessary.

Calibrated cham-ber

Use a calibrated chamber: a calibrated cylindrical ionization chamber with acavity volume of at least 0.125 cm3 but not more than 0.6 cm3. The effectivepoint of measurement shall be determined based on valid international dos-imetry standards (e.g. IAEA TRS-398) and the corresponding recommenda-tion of the detector provider.

Ionization cham-ber

Use an ionization chamber: A cylindrical ionization chamber with a cavity vol-ume of 0.125 cm3 or smaller. The effective point of measurement shall bedetermined based on valid international dosimetry standards (e.g. IAEATRS-398) and the corresponding recommendation of the detector provider.

High-resolutiondetector

Use a very small detector for high-resolution profile measurements and dos-imetry of small fields. Brainlab recommends the use of an unshielded stereo-tactic diode or a single crystal diamond detector.

Small build-upcap

Use a small build-up cap made of brass or material of similar mass densityfor measurements in air. The thickness of build-up cap:

dcap10E

3ρcap--------------,≈ dcap in mm, E in MV, ρcap in g/cm3( ),

for brass: dbrass 6 MV( ) 2.4 mm, dbrass 15 MV( ) 5.9 mm.≈≈



7.2 Coordinate Systems

Background

The measurements described in this document are based on the coordinate system illustratedbelow.• All length units are given in mm.• For all measurements, the collimator angle and the gantry angle must be both set to 0°.• Consult the specific checklist for your MLC (see page 117).

Coordinate System

①

②

Figure 26

No. Component

① Measurements in air

② Measurements in water

Understanding the Coordinate System

The coordinate system corresponds to the fixed system of IEC 1217, but rotated by 180° aroundthe X-axis, i.e. the Y and Z point in opposite directions.

Measurement Explanation

Air

For all measurements in air, the origin of the coordinate system is not locatedin the isocenter, but in the nominal photon source point in the target, i.e., at adistance of SID away from the isocenter as indicated above. This means thatZ coordinates for all measurements in air (especially the Z profiles) must bespecified as a distance to the source (focus) point in millimeters.

Water

For measurements in water, the coordinate system is slightly different. Herethe origin is located at the surface of the water phantom, i.e. depth dosecurves are measured as dose per depth in water in millimeters. The X, Y andZ directions remain the same. These directions are independent of the mount-ing direction of the MLC. In other words, MLCs with leaves moving in X direc-tion and MLCs with leaves moving in Y direction are possible.

Coordinate Systems


7.3 Data Correction

Background

A limited level of data correction is allowed in order to eliminate small errors during measurementdata acquisition. However, such corrections must be approached with caution. It is always betterto avoid corrections by measuring data that does not need to be modified.For example, to take central axis deviations caused by measurement errors into account, it isuseful to shift the profiles by the corresponding off-axis distance.It might also be useful to symmetrize the measured profiles because the fitting algorithm assumessymmetric profiles. It is always better if the accelerator can produce symmetric (or almostsymmetric) and flat profiles.Symmetrization should not be performed before centering the profiles (accounting for central axisdeviations).

Data Format

All measured data (in water and air) must be entered into the Monte Carlo Raw Data mode ofPhysics Administration provided by Brainlab. The resulting files (with the file extension*.xmcdat) containing the complete set of measurements per photon energy must be sent [email protected].




7.4 Beam Data Measurements in Air

Ionization Chambers

Use an ionization chamber with a brass build up cap (or similar material) to measure the X, Y andZ profiles in air for all photon energies and different field sizes. The field sizes for your MLC aredocumented in the corresponding checklist.

Build-up Cap

The main purpose of the brass build-up cap during the in-air scans is to remove electrons createdin the linac head from the measurement signal. Therefore, the cap thickness must be larger thanthe maximum range of these electrons.On the other hand, the spatial resolution of the in-air profile scans reduces and volume averagingartifacts influence the small field in-air output factor measurements due to larger wall thicknesses.Therefore, the formula on page 111 has to be considered as a compromise between bothrequirements. If a build-up cap according to this recommendation is not available, a cap withlarger thickness may be used because removing electron contamination is more important thanspatial resolution.

MLC and Jaw Field Shapes

If possible, only the jaws should collimate the fields (i.e., the MLC leaves should be fullyretracted). If this is not possible (e.g., for Elekta MLCs), use the MLC like a pair of jaws. For alinac with a MLC only (no jaws; e.g., MHI), use the MLC for field collimation.Always provide the correct MLC and jaw field shapes for each field size, which can be specifiedusing the MLC and jaw field size settings in the Monte Carlo Raw Data file (*.xmcdat). Set theMLC or jaw field size to 400 mm, if the corresponding X or Y device is not available or fullyretracted.The following data must be measured:• Z profile (depth dose) at the central axis (X = Y = 0) from about Z = 850 to Z = 1150 (Z = 0:

corresponds to the nominal focus of the photon source)• 3 X profiles for Y = 0 and Z = 850, Z = 1000, Z = 1150• 3 Y profiles for X = 0 and Z = 850, Z = 1000, Z = 1150• In-air output factors at Z = 1000 for all field sizes normalized by one of the fields, usually the

100 x 100 mm² field

Beam Data without Measurements in Air

In general, commissioning of the Monte Carlo (MC) dose calculation algorithm requires dosemeasurements (output factors, cross profiles and depth dose curves) for different square andrectangular field sizes in air using an ionization chamber with a brass build-up cap. Thesemeasurements are required by Brainlab to adjust certain geometric parameters of the virtual linachead model. However, an analysis of the internal database of MC customer measurementsdemonstrated only minor in-air data variation, especially for Novalis Tx, TrueBeam and TrueBeamSTx linacs with identical energy and beam mode (i.e., standard (STD) mode, stereotacticradiosurgery (SRS) mode or flattening filter free (FFF) mode). Therefore, in-air data acquisitionmay be omitted for the following linac types, energies [MV] and flattening filter modes:

Novalis Tx (HD 120)- STD: 6, 10- SRS: 6

TrueBeam (Millennium 120)- STD: 6, 10

TrueBeam STx (HD 120)- STD: 4, 6, 10, 15- FFF: 6, 10

Beam Data Measurements in Air


For all other combinations of linac type, energy and flattening filter mode, all measurements in airmust be performed as customer data sets are insufficient to analyze the measurement variations.If MC beam data without in-air measurements are submitted to Brainlab, an accurate virtual linachead model may not be able to be created. For example, if there is no agreement between themeasured data in water and the results of the verification calculations. In this case, the in-air datamust be acquired afterwards. To avoid such a situation, Brainlab recommends measuring MC datain air for all combinations of linac type, energy and flattening filter mode.To input MC measurement data and create the MC Raw Data file with or without in-air data, referto the Physics Administration Software User Guide.



7.5 Beam Data Measurements in Water

MLC and Jaw Field Shapes

Always provide the correct MLC and jaw field shapes for each field size, which can be specifiedusing the MLC and jaw field size settings in the Monte Carlo Raw Data file (*.xmcdat). Set theMLC or jaw field size to 400 mm, if the corresponding X or Y device is not available or fullyretracted.

Absolute Dose

You must measure the absolute dose in Gray per monitor unit (Gy/MU) for the 100 x 100 mm² fieldsize for both SSD = 900 mm and SSD = 1000 mm in 100 mm reference depth of the waterphantom. Here, an ionization chamber calibrated for absolute dose measurements must be used.Both the jaws and the MLC must collimate the field.

Measurements in Water (SSD=1000 mm)

X, Y and Z (depth dose) profiles in water are required for all photon energies and for the field size100 x 100 mm2 (source-surface distance of the water phantom: SSD = 1000 mm). Both the jawsand the MLC must collimate the fields:• Z profile (depth dose) at the central axis (X = Y = 0), Z = 0: corresponds to the surface of the

water phantom• 3 X profiles for Y = 0 and depth of maximum dose, Z = 100, Z = 200• 3 Y profiles for X = 0 and depth of maximum dose, Z = 100, Z = 200

The X and Y profiles must be measured using detectors small enough for the corresponding fieldsize and to reproduce the penumbra correctly. The central axis depth dose curve shall bemeasured with an ionization chamber or a diode detector. It is used to determine the photonenergy distribution and the amount of electron contamination.

Measurements in Water (SSD=900 mm)

X, Y and Z (depth dose) profiles in water are required for all photon energies and different fieldsizes (source-surface distance of the water phantom: SSD = 900 mm). The field sizes for yourMLC are documented in the corresponding checklist (see page 117). Both the jaws and the MLCmust collimate the fields:• Z profile (depth dose) at the central axis (X = Y = 0), Z = 0: corresponds to the surface of the

water phantom• 3 X profiles for Y = 0 and depth of maximum dose, Z = 100, Z = 200• 3 Y profiles for X = 0 and depth of maximum dose, Z = 100, Z = 200• Output factors at Z = 100 for all field sizes normalized by the 100 x 100 field at SSD = 900 mm

All PDDs, X profiles, Y profiles and the output factors shall be measured using detectors smallenough for the corresponding field size and to reproduce the penumbra correctly (see page 21).Large field PDDs and output factors shall be measured using an ionization chamber of medium orlarge volume (0.1 cm3 – 0.6 cm3) to avoid the effect of energy response variations (see sectionIII.D.5 and Fig. 1 of the AAPM TG-106 Report (Das et al 2008)).

Ensuring Accuracy

To obtain an accurate absorbed dose to water measurement, you must apply a number ofcorrections to the dosimeter reading (e.g., beam quality (linac energy), pressure, temperature andpolarity). You must consult the documentation provided with your dosimetric equipment and thenational standards applicable in your country.

Beam Data Measurements in Water


8 MONTE CARLO: BEAMDATA CHECKLISTS

8.1 Beam Data for Elekta Agility

Guard Leaves

Always specify using the MLC and jaw field size settings in the Monte Carlo Raw Data file(*.xmcdat), whether guard leaves were used or not. Guard leaves are two extra open leaf pairs oneach side under the jaws adjacent to and in the same physical position as the last in-field leaf pair.For example, a 30x30 mm² field with guard leaves is specified by a 30x50 mm² MLC field size anda 400x30 mm² jaw field size and a 30x30 mm² field without guard leaves is specified by a 30x30mm² MLC field size and a 400x30 mm² jaw field size. Always set the X jaw field size to 400 mmbecause X jaws are not available.

Checklist

Task No. of Measure-ments

Equipment See Also Done

CAX profiles in airMLC: with guard leaves;Field sizes: 20 x 20, 30 x 30, 60 x 60, 100 x 100,200 x 200, 400 x 400, 50 x 400, 400 x 50Z = 850 … 1150

8Ionization cham-ber with build-upcap

Page 114 ☐

X profiles in airMLC: with guard leaves;Field sizes: 20 x 20, 30 x 30, 60 x 60, 100 x 100,200 x 200, 400 x 400, 50 x 400, 400 x 50;Y = 0; Z = 850, 1000, 1150


☐

Y profiles in airMLC: with guard leaves;Field sizes: 20 x 20, 30 x 30, 60 x 60, 100 x 100,200 x 200, 400 x 400, 50 x 400, 400 x 50;X = 0; Z = 850, 1000, 1150


☐

Output factors in airMLC: with guard leaves;Field sizes: 20 x 20, 30 x 30, 60 x 60, 100 x 100,200 x 200, 400 x 400, 50 x 400, 400 x 50;X = 0; Y = 0; Z = 1000


☐

Absolute dose in Gy per MUMLC field size: 100 x 120 or 100 x 100;Y jaws opening: 100;SSD = 900; X = 0; Y = 0; Z = 100

1 Calibrated cham-ber Page 116 ☐

MONTE CARLO: BEAM DATA CHECKLISTS




Absolute dose in Gy per MUMLC field size: 100 x 120 or 100 x 100;Y jaws opening: 100;SSD = 1000; X = 0; Y = 0; Z = 100

1 Calibrated cham-ber ☐

CAX PDD in waterMLC field size: 100 x 120 or 100 x 100;Y jaws opening: 100;SSD = 1000

1 Ionization cham-ber Page 116 ☐

X profiles in waterMLC field size: 100 x 120 or 100 x 100;Y jaws opening: 100;SSD = 1000; Y = 0; Z = Zmax, 100, 200

3 High-resolutiondetector ☐

Y profiles in waterMLC field size: 100 x 120 or 100 x 100;Y jaws opening: 100;SSD = 1000; X = 0; Z = Zmax, 100, 200


CAX PDDs in waterMLC (X x Y) and jaw (Y) field sizes:6 x 30 (6) with collimator angle 0°,6 x 30 (6) with collimator angle 90°,10 x 30 (10) or 10 x 10 (10),30 x 50 (30) or 30 x 30 (30),60 x 80 (60) or 60 x 60 (60),100 x 120 (100) or 100 x 100 (100),200 x 220 (200) or 200 x 200 (200),400 x 400 (400),50 x 400 (400),400 x 70 (50) or 400 x 50 (50),SSD = 900

10Ionization cham-ber and high-res-olution detector

Page 116 ☐

X profiles in waterMLC (X x Y) and jaw (Y) field sizes:6 x 30 (6) with collimator angle 0°,6 x 30 (6) with collimator angle 90°,10 x 30 (10) or 10 x 10 (10),30 x 50 (30) or 30 x 30 (30),60 x 80 (60) or 60 x 60 (60),100 x 120 (100) or 100 x 100 (100),200 x 220 (200) or 200 x 200 (200),400 x 400 (400),50 x 400 (400),400 x 70 (50) or 400 x 50 (50),SSD = 900; Y = 0; Z = Zmax, 100, 200


Beam Data for Elekta Agility




Y profiles in waterMLC (X x Y) and jaw (Y) field sizes:6 x 30 (6) with collimator angle 0°,6 x 30 (6) with collimator angle 90°,10 x 30 (10) or 10 x 10 (10),30 x 50 (30) or 30 x 30 (30),60 x 80 (60) or 60 x 60 (60),100 x 120 (100) or 100 x 100 (100),200 x 220 (200) or 200 x 200 (200),400 x 400 (400),50 x 400 (400),400 x 70 (50) or 400 x 50 (50),SSD = 900; X = 0; Z = Zmax, 100, 200


Output factors in waterMLC (X x Y) and jaw (Y) field sizes:6 x 30 (6) with collimator angle 0°,6 x 30 (6) with collimator angle 90°,10 x 30 (10) or 10 x 10 (10),30 x 50 (30) or 30 x 30 (30),60 x 80 (60) or 60 x 60 (60),100 x 120 (100) or 100 x 100 (100),200 x 220 (200) or 200 x 200 (200),400 x 400 (400),50 x 400 (400),400 x 70 (50) or 400 x 50 (50),SSD = 900; X = 0; Y = 0, Z = 100


☐

How to Measure with Different Collimator Angles for Small Field Sizes

To support the Monte Carlo dose calculation for field sizes down to 6 x 6 mm², output factors andprofiles for this field size must always be measured with guard leaves as well as with collimatorangles 0° and 90°.

Step

1. Set the MLC field size to 6 x 30 mm² (with guard leaves) and the Y jaw field size to 6 mm.

2. Measure the output factor, PDD and the cross profiles using collimator angle 0°.

3. Rotate the collimator by 90° while leaving the MLC and jaw setting unchanged.

4. Measure the output factor, PDD and cross profiles again.This is necessary to model the oval shape of the primary photon source within the brems-strahlung target.

The shape of the primary photon source in combination with the guard leaves behavior isresponsible for the change of the dose output of very small fields when the collimator is rotated.



8.2 Beam Data for MHI MLC 60

Checklist



CAX profiles in airMLC field sizes: 20 x 20, 30 x 30, 50 x 50, 70 x 70,100 x 100, 150 x 150, 50 x 150, 150 x 50Z = 850 ... 1150


Page 114 ☐

X profiles in airMLC field sizes: 20 x 20, 30 x 30, 50 x 50, 70 x 70,100 x 100, 150 x 150, 50 x 150, 150 x 50Y = 0; Z = 850, 1000, 1150


☐

Y profiles in airMLC field sizes: 20 x 20, 30 x 30, 50 x 50, 70 x 70,100 x 100, 150 x 150, 50 x 150, 150 x 50X = 0; Z = 850, 1000, 1150


☐

Output factors in airMLC field sizes: 20 x 20, 30 x 30, 50 x 50, 70 x 70,100 x 100, 150 x 150, 50 x 150, 150 x 50X = 0; Y = 0; Z = 1000


☐

Absolute dose in Gy per MUMLC field size: 100 x 100;SSD = 900; X = 0; Y = 0; Z = 100


Absolute dose in Gy per MUMLC field size: 100 x 100;SSD = 1000; X = 0; Y = 0; Z = 100


CAX PDD in waterMLC field size: 100 x 100;SSD = 1000


X profiles in waterMLC field size: 100 x 100;SSD = 1000; Y = 0; Z = Zmax, 100, 200


Y profiles in waterMLC field size: 100 x 100;SSD = 1000; X = 0; Z = Zmax, 100, 200


CAX PDDs in waterMLC field sizes: 10 x 10, 30 x 30, 50 x 50, 70 x 70,100 x 100, 150 x 150, 50 x 150, 150 x 50;SSD = 900


Page 116 ☐

X profiles in waterMLC field sizes: 10 x 10, 30 x 30, 50 x 50, 70 x 70,100 x 100, 150 x 150, 50 x 150, 150 x 50;SSD = 900; Y = 0; Z = Zmax, 100, 200


Y profiles in waterMLC field sizes: 10 x 10, 30 x 30, 50 x 50, 70 x 70,100 x 100, 150 x 150, 50 x 150, 150 x 50;SSD = 900; X = 0; Z = Zmax, 100, 200


Beam Data for MHI MLC 60




Output factors in waterMLC field sizes: 10 x 10, 30 x 30, 50 x 50, 70 x 70,100 x 100, 150 x 150, 50 x 150, 150 x 50;SSD = 900; X = 0; Y = 0, Z = 100


☐



8.3 Beam Data for Novalis/Brainlab m3

Checklist



CAX profiles in airMLC field sizes: fully retractedJaw field sizes: 18 x 18, 24 x 24, 42 x 42, 60 x 60,80 x 80, 98 x 98, 24 x 98, 98 x 24Z = 850 ... 1150


Page 114 ☐

X profiles in airMLC field sizes: fully retractedJaw field sizes: 18 x 18, 24 x 24, 42 x 42, 60 x 60,80 x 80, 98 x 98, 24 x 98, 98 x 24Y = 0; Z = 850, 1000, 1150


☐

Y profiles in airMLC field sizes: fully retractedJaw field sizes: 18 x 18, 24 x 24, 42 x 42, 60 x 60,80 x 80, 98 x 98, 24 x 98, 98 x 24X = 0; Z = 850, 1000, 1150


☐

Output factors in airMLC field sizes: fully retractedJaw field sizes: 18 x 18, 24 x 24, 42 x 42, 60 x 60,80 x 80, 98 x 98, 24 x 98, 98 x 24X = 0; Y = 0; Z = 1000


☐

Absolute dose in Gy per MUMLC field size: 100 x 100;Jaw field size: 98 x 98SSD = 1000; X = 0; Y = 0; Z = 100


Absolute dose in Gy per MUMLC field size: 100 x 100;Jaw field size: 98 x 98SSD = 900; X = 0; Y = 0; Z = 100


CAX PDD in waterMLC field size: 100 x 100;Jaw field size: 98 x 98SSD = 1000


X profiles in waterMLC field size: 100 x 100;Jaw field size: 98 x 98SSD = 1000; Y = 0; Z = Zmax, 100, 200


Y profiles in waterMLC field size: 100 x 100;Jaw field size: 98 x 98SSD = 1000; X = 0; Z = Zmax, 100, 200


Beam Data for Novalis/Brainlab m3




CAX PDDs in waterMLC (jaw) field sizes:6 x 6 (8 x 8)12 x 12 (14 x 14),18 x 18 (20 x 20),24 x 24 (26 x 26),42 x 42 (44 x 44),60 x 60 (60 x 60),100 x 100 (98 x 98),24 x 100 (26 x 98),100 x 24 (98 x 26),SSD = 900


Page 116 ☐

X profiles in waterMLC (jaw) field sizes:6 x 6 (8 x 8)12 x 12 (14 x 14),18 x 18 (20 x 20),24 x 24 (26 x 26),42 x 42 (44 x 44),60 x 60 (60 x 60),100 x 100 (98 x 98),24 x 100 (26 x 98),100 x 24 (98 x 26),SSD = 900; Y = 0; Z = Zmax, 100, 200


Y profiles in waterMLC (jaw) field sizes:6 x 6 (8 x 8)12 x 12 (14 x 14),18 x 18 (20 x 20),24 x 24 (26 x 26),42 x 42 (44 x 44),60 x 60 (60 x 60),100 x 100 (98 x 98),24 x 100 (26 x 98),100 x 24 (98 x 26),SSD = 900; X = 0; Z = Zmax, 100, 200


Output factors in waterMLC (jaw) field sizes:6 x 6 (8 x 8)12 x 12 (14 x 14),18 x 18 (20 x 20),24 x 24 (26 x 26),42 x 42 (44 x 44),60 x 60 (60 x 60),100 x 100 (98 x 98),24 x 100 (26 x 98),100 x 24 (98 x 26),SSD = 900; X = 0; Y = 0, Z = 100


☐



8.4 Beam Data for Varian HD120 (SRS FlatteningFilter)

Checklist




CAX profiles in airMLC field sizes: fully retractedJaw field sizes: 20 x 20, 40 x 40, 60 x 60, 80 x 80,100 x 100, 150 x 150, 50 x 150, 150 x 50Z = 850 ... 1150


Page 114 ☐



☐



☐



☐

Absolute dose in Gy per MUMLC and jaw field size: 100 x 100;SSD = 900; X = 0; Y = 0; Z = 100




CAX PDD in waterMLC and jaw field size: 100 x 100;SSD = 1000


X profiles in waterMLC and jaw field size: 100 x 100;SSD = 1000; Y = 0; Z = Zmax, 100, 200


Y profiles in waterMLC and jaw field size: 100 x 100;SSD = 1000; X = 0; Z = Zmax, 100, 200


Beam Data for Varian HD120 (SRS Flattening Filter)




CAX PDDs in waterMLC (jaw) field sizes:5 x 5 (8 x 8)10 x 10 (12 x 12),20 x 20 (22 x 22),40 x 40 (42 x 42),60 x 60 (60 x 60),100 x 100 (100 x 100),150 x 150 (150 x 150),50 x 150 (50 x 150),150 x 50 (150 x 50);SSD = 900


Page 116 ☐

X profiles in waterMLC (jaw) field sizes:5 x 5 (8 x 8)10 x 10 (12 x 12),20 x 20 (22 x 22),40 x 40 (42 x 42),60 x 60 (60 x 60),100 x 100 (100 x 100),150 x 150 (150 x 150),50 x 150 (50 x 150),150 x 50 (150 x 50);SSD = 900; Y = 0; Z = Zmax, 100, 200


Y profiles in waterMLC (jaw) field sizes:5 x 5 (8 x 8)10 x 10 (12 x 12),20 x 20 (22 x 22),40 x 40 (42 x 42),60 x 60 (60 x 60),100 x 100 (100 x 100),150 x 150 (150 x 150),50 x 150 (50 x 150),150 x 50 (150 x 50),SSD = 900; X = 0; Z = Zmax, 100, 200


Output factors in waterMLC (jaw) field sizes:5 x 5 (8 x 8)10 x 10 (12 x 12),20 x 20 (22 x 22),40 x 40 (42 x 42),60 x 60 (60 x 60),100 x 100 (100 x 100),150 x 150 (150 x 150),50 x 150 (50 x 150),150 x 50 (150 x 50);SSD = 900; X = 0; Y = 0, Z = 100


☐



8.5 Beam Data for Varian HD120 (Standard Irradiationand Flattening Filter Free Mode)

Checklist



CAX profiles in airMLC field sizes:fully retractedJaw field sizes: 20 x 20, 40 x 40, 60 x 60, 100 x 100,150 x 150, 220 x 220, 50 x 220, 250 x 50Z = 850 ... 1150


Page 114 ☐



☐



☐



☐











Beam Data for Varian HD120 (Standard Irradiation and Flattening Filter Free Mode)




CAX PDDs in waterMLC (jaw) field sizes:5 x 5 (8 x 8)10 x 10 (12 x 12),20 x 20 (22 x 22),40 x 40 (42 x 42),60 x 60 (60 x 60),100 x 100 (100 x 100),220 x 220 (220 x 220),50 x 220 (50 x 220),250 x 50 (250 x 50);SSD = 900


Page 116 ☐

X profiles in waterMLC (jaw) field sizes:5 x 5 (8 x 8)10 x 10 (12 x 12),20 x 20 (22 x 22),40 x 40 (42 x 42),60 x 60 (60 x 60),100 x 100 (100 x 100),220 x 220 (220 x 220),50 x 220 (50 x 220),250 x 50 (250 x 50);SSD = 900; Y = 0; Z = Zmax, 100, 200


Y profiles in waterMLC (jaw) field sizes:5 x 5 (8 x 8)10 x 10 (12 x 12),20 x 20 (22 x 22),40 x 40 (42 x 42),60 x 60 (60 x 60),100 x 100 (100 x 100),220 x 220 (220 x 220),50 x 220 (50 x 220),250 x 50 (250 x 50);SSD = 900; X = 0; Z = Zmax, 100, 200


Output factors in waterMLC (jaw) field sizes:5 x 5 (8 x 8)10 x 10 (12 x 12),20 x 20 (22 x 22),40 x 40 (42 x 42),60 x 60 (60 x 60),100 x 100 (100 x 100),220 x 220 (220 x 220),50 x 220 (50 x 220),250 x 50 (250 x 50),SSD = 900; X = 0; Y = 0, Z = 100


☐



8.6 Beam Data for Varian 120 (SRS Flattening Filter)

Checklist

NOTE: SRS mode refers to the SRS-mode flattening filter of Novalis TX and Varian Trilogy linacs.This SRS mode uses 6 MV photon beams and a high dose rate of 1000 MU/min in combinationwith a limited maximum field size of 150 x150 mm². If the linac is in SRS mode, it uses a differentflattening filter. Hence, beam data for Standard and SRS mode are slightly different.



CAX profiles in airMLC field size:400 x 400;Jaw field sizes: 20 x 20, 40 x 40, 60 x 60, 80 x 80,100 x 100, 150 x 150, 50 x 150, 150 x 50Z = 850 ... 1150


Page 114 ☐

X profiles in airMLC field size:400 x 400;Jaw field sizes: 20 x 20, 40 x 40, 60 x 60, 80 x 80,100 x 100, 150 x 150, 50 x 150, 150 x 50Y = 0; Z = 850, 1000, 1150


☐

Y profiles in airMLC field size:400 x 400;Jaw field sizes: 20 x 20, 40 x 40, 60 x 60, 80 x 80,100 x 100, 150 x 150, 50 x 150, 150 x 50X = 0; Z = 850, 1000, 1150


☐

Output factors in airMLC field size:400 x 400;Jaw field sizes: 20 x 20, 40 x 40, 60 x 60, 80 x 80,100 x 100, 150 x 150, 50 x 150, 150 x 50X = 0; Y = 0; Z = 1000


☐











Beam Data for Varian 120 (SRS Flattening Filter)




CAX PDDs in waterMLC (jaw) field sizes:10 x 10 (12 x 12),20 x 20 (22 x 22),40 x 40 (42 x 42),60 x 60 (60 x 60),100 x 100 (100 x 100),150 x 150 (150 x 150),50 x 150 (50 x 150),150 x 50 (150 x 50);SSD = 900


Page 116 ☐

X profiles in waterMLC (jaw) field sizes:10 x 10 (12 x 12),20 x 20 (22 x 22),40 x 40 (42 x 42),60 x 60 (60 x 60),100 x 100 (100 x 100),150 x 150 (150 x 150),50 x 150 (50 x 150),150 x 50 (150 x 50);SSD = 900; Y = 0; Z = Zmax, 100, 200


Y profiles in waterMLC (jaw) field sizes:10 x 10 (12 x 12),20 x 20 (22 x 22),40 x 40 (42 x 42),60 x 60 (60 x 60),100 x 100 (100 x 100),150 x 150 (150 x 150),50 x 150 (50 x 150),150 x 50 (150 x 50);SSD = 900; X = 0; Z = Zmax, 100, 200


Output factors in waterMLC (jaw) field sizes:10 x 10 (12 x 12),20 x 20 (22 x 22),40 x 40 (42 x 42),60 x 60 (60 x 60),100 x 100 (100 x 100),150 x 150 (150 x 150),50 x 150 (50 x 150),150 x 50 (150 x 50);SSD = 900; X = 0; Y = 0, Z = 100


☐



8.7 Beam Data for Varian 120 (Standard Irradiationand Flattening Filter Free Mode)

Checklist



CAX profiles in airMLC field size: 400 x 400;Jaw field sizes: 20 x 20, 40 x 40, 60 x 60, 100 x 100,200 x 200, 400 x 400, 50 x 400, 400 x 50Z = 850 ... 1150


Page 114 ☐

X profiles in airMLC field size: 400 x 400;Jaw field sizes: 20 x 20, 40 x 40, 60 x 60, 100 x 100,200 x 200, 400 x 400, 50 x 400, 400 x 50Y = 0; Z = 850, 1000, 1150


☐

Y profiles in airMLC field size: 400 x 400;Jaw field sizes: 20 x 20, 40 x 40, 60 x 60, 100 x 100,200 x 200, 400 x 400, 50 x 400, 400 x 50X = 0; Z = 850, 1000, 1150


☐

Output factors in airMLC field size: 400 x 400;Jaw field sizes: 20 x 20, 40 x 40, 60 x 60, 100 x 100,200 x 200, 400 x 400, 50 x 400, 400 x 50X = 0; Y = 0; Z = 1000


☐











Beam Data for Varian 120 (Standard Irradiation and Flattening Filter Free Mode)




CAX PDDs in waterMLC (jaw) field sizes:10 x 10 (12 x 12),20 x 20 (22 x 22),40 x 40 (42 x 42),60 x 60 (60 x 60),100 x 100 (100 x 100),300 x 300 (300 x 300),50 x 300 (50 x 300),300 x 50 (300 x 50);SSD = 900


Page 116 ☐

X profiles in waterMLC (jaw) field sizes:10 x 10 (12 x 12),20 x 20 (22 x 22),40 x 40 (42 x 42),60 x 60 (60 x 60),100 x 100 (100 x 100),300 x 300 (300 x 300),50 x 300 (50 x 300),300 x 50 (300 x 50);SSD = 900; Y = 0; Z = Zmax, 100, 200


Y profiles in waterMLC (jaw) field sizes:10 x 10 (12 x 12),20 x 20 (22 x 22),40 x 40 (42 x 42),60 x 60 (60 x 60),100 x 100 (100 x 100),300 x 300 (300 x 300),50 x 300 (50 x 300),300 x 50 (300 x 50),SSD = 900; X = 0; Z = Zmax, 100, 200


Output factors in waterMLC (jaw) field sizes:10 x 10 (12 x 12),20 x 20 (22 x 22),40 x 40 (42 x 42),60 x 60 (60 x 60),100 x 100 (100 x 100),300 x 300 (300 x 300),50 x 300 (50 x 300),300 x 50 (300 x 50),SSD = 900; X = 0; Y = 0, Z = 100


☐



Beam Data for Varian 120 (Standard Irradiation and Flattening Filter Free Mode)


9 DYNAMIC DELIVERY9.1 Introduction

Background

Brainlab treatment planning applications support either dynamic arcs or Volumetric Modulated ArcTherapy (VMAT) arcs. For dynamic arcs, the leaves conform to the PTV shape at each controlpoint. The dose rate and gantry speed are constant. In contrast, for VMAT arcs, the leaf positionsare independent of the PTV shape. Their movement between control points is mainly restricted bythe maximum leaf speed.Depending on the capabilities of the system (linac and MLC), the dose rate and gantry speed canbe both constant (e.g., linacs equipped with m3 MLC), both variable (e.g., most modern linacs) orthe dose rate is variable while the gantry speed is constant (e.g., older linacs).

DYNAMIC DELIVERY


9.2 Deliverability of Arcs

Leaf Sequencing Overview

Brainlab's radiotherapy treatment planning software creates a leaf movement pattern for dynamicarc fields and VMAT fields using the built-in leaf sequencing algorithm. In patterns of this kind, theleaf positions are defined at a number of control points as a function of the cumulative fractionaldose that has been delivered so far.

Correct Leaf Movement and Segment Dose During Dose Delivery

In order to correctly deliver the computed leaf movement pattern, the system (linac and MLC)must follow the exact leaf pattern calculated as a function of the cumulative fractional dose, andalso irradiate the required fractional dose for each segment (segment dose) accurately.

Leaf Movement To Home Position

Leaf pairs that are not used at the current control point are moved to their home position, wherethey may be covered by the jaws. This is typically as far off-center as possible. If a leaf pair is notused in any control point, it stays at the home position for the whole duration of the arc. If a leafpair is used in at least one control point, it is temporarily moved to the home position whenever itis not used. However, the distance to the home position might be too large to travel there andback until the leaf pair is used again. In this case, the leaf pair is parked near the border of thePTV.

Dose Deviations

Deviations between the planned and delivered dose may occur if the system either cannotposition the leaves correctly according to the prescribed fractional dose, or does not irradiate therequired fractional dose for a given VMAT segment. Examples are provided below of scenariosthat may lead to significant deviations in dose delivery to the patient:• General limitations in the leaf positioning accuracy of the MLC.• During VMAT delivery, the system does not reach the required positions for a certain delivered

fractional dose. This is typically the case if the maximum leaf speed of the MLC is exceededand/or a high leaf tolerance value is used (see page 136). Make sure to enter the correctmachine constraints into Physics Administration when creating the machine profile. Thisenables the VMAT leaf sequencer to create plans that can be successfully delivered or issuecorrect warnings if the machine constraints are violated.

• Among other parameters, the VMAT leaf sequencer uses the maximum gantry speed and themaximum dose rate of the linac to derive various machine constraints. Setting an incorrectmaximum gantry speed or maximum dose rate in Physics Administration when creating themachine profile may lead to undeliverable plans. In this case, the leaf sequencer cannot issuecorrect warnings if the machine constraints are violated.

• Due to the response time of the linac/MLC controller system and/or random irregularities in theradiation stability, the delivered fractional dose per VMAT segment can deviate from theplanned fractional dose. This is particularly true in the case of small fractional doses persegment. For example, if the dose per segment is 2 MU and the linac delivers this dose towithin ±1 MU, the potential dose error per segment may be as large as 50%. A higher dose ratecan amplify these dose deviations.

• The dose calculation may be inaccurate if the VMAT leaf shapes are excessively complex orfragmented. This is usually connected with an unusually high total number of MUs. Consider todecrease the modulation for such plans if this is supported by your Brainlab treatment planningapplication. Such plans require thorough patient-specific QA.

Information about machine QA and patient-specific QA for VMAT can be found for example in thefollowing papers: Clifton Ling 2008, Bedford 2009, Masi 2011, Van Esch 2011 and Wang 2013.

Deliverability of Arcs


Safety Notes

The scenarios described here are merely examples of possible limitations that can lead tosignificant deviations between the planned and delivered dose. These scenarios do not provide acomplete description of potential problems. It should be carefully verified whether complex VMATplans, for example with low segment doses, lead to acceptable treatment results.

WarningThe maximum allowed jaw speed is not considered by the planning system for DynamicArc and VMAT treatments using Elekta Agility. In rare cases, this may lead to inaccuratedose delivery. Make sure your routine quality assurance tests are able to detect theseinaccuracies. According to tests, the linac controller reduces the dose rate in cases wherethe jaw speed would be exceeded. This may lead to a slightly longer treatment deliverytime. In other cases, treatment plans may possibly be rejected by the delivery system dueto this fact.

WarningUsing a treatment delivery system with extreme parameters (e.g., high dose rate, low MUper beam or high leaf speed) may result in inaccurate delivery of the planned treatmentdose. It is the physicist’s responsibility to ensure correct delivery of the planned treatmentdose to the patient. This includes responsibility for the choice of appropriate treatmentparameters.

DYNAMIC DELIVERY


9.3 Leaf Tolerance

Background

The leaf sequencing patterns for arc treatments using a Varian linac are defined by DMLC files.These files contain the planned leaf positions as a function of the cumulative fractional dose(relative cumulative dose value) for certain segments. These files are sent to the MLC controllerby the R&V system.Under normal conditions, i.e. as long as the maximum leaf speed is not exceeded, the MLCcontroller in combination with the linac can follow the planned leaf sequencing pattern with anacceptable accuracy (assuming a linear dependency of the leaf positions between consecutivesegments). In order to ensure accurate delivery, the MLC workstation samples the actual leafpositions and the cumulative fractional dose delivered by the linac so far at a particular repetitionrate (the current sampling time of the Varian MLC controller is about 55 ms). These results arethen compared to the values defined by the leaf sequence. If the maximum leaf speed isexceeded by some leaves at certain points of the delivery, the actual leaf positions deviate fromthe planned ones. If this deviation exceeds the tolerance value defined in the DMLC file, the MLCworkstation sends a beam hold signal to the linac in order to stop irradiation.

Requirements for Leaf Position Accuracy

Choosing a large tolerance value to allow large deviations in leaf position may result in significantdifferences between the delivered and planned dose. On the other hand, choosing very smalltolerance values may result in a large number of beam holds, which can also increase thedeviation between the planned and the delivered dose. Further information is provided in thepublication by Hernandez (2015).In order to verify the machine constraints stored in the machine profile and the chosen tolerancevalue for arc delivery, phantom treatments must be performed with various plans of differentcomplexity and with absolute dose measurements for the entire irradiated volume. Thesemeasurements must then be compared to the planned/calculated dose distribution.The leaf position and dose errors logged by the MLC workstation (DynaLog file) may also be usedto isolate potential problems in plan delivery. The leaf tolerance value can be adjusted in the R&Vsystem. Further information on adjusting this parameter is provided in the documentation for yourR&V system.

Safety Notes

Prior to treatment plan verification using a phantom, an analysis of the leaf position errors that arelogged by the MLC workstation (DynaLog file) may be useful in order to detect potential deviationsin plan delivery. For further information on handling and evaluating the data, refer to theappropriate documentation as required.

Leaf Tolerance


9.4 Dynamic Leaf Shift for Modulated Treatments

Background

For conformal beams, and for static and dynamic conformal arcs the static leaf shift is used tocorrect for the transmission of radiation through the rounded leaf tips of most MLCs. Thus, thestatic leaf shift connects the physical leaf position with the radiological leaf position.In modulated treatments such as dynamic IMRT or VMAT, leaves often cover parts of the PTV.This increases the contribution of dose leakage to the total dose. In order to model this, thedynamic leaf shift can be used, see page 62. The model works quite well for sliding window IMRTplans because in this case the leaves travel across the PTV exactly once per beam.In VMAT arcs, the situation is much more complex compared to dynamic IMRT. It was suggestedthat for every VMAT plan there is a different optimal dynamic leaf shift, see Kielar 2012 and Yao2015. However, a single, optimized value for the radiologic leaf shift may yield sufficientagreements between the planned and delivered dose.The VMAT algorithm of the Brainlab RT Elements allows to use either the static leaf shift or thedynamic leaf shift during treatment planning. This option is set in the Machine Profile and can bechanged using Physics Administration. For details, refer to the Physics AdministrationSoftware User Guide.

Safety Notes

It is the responsibility of the physicist to check which static or dynamic leaf shift is optimal on theirmachine. This might differ between IMRT and VMAT plans.

DYNAMIC DELIVERY


Dynamic Leaf Shift for Modulated Treatments


10 QUALITY ASSURANCE10.1 Introduction to Quality Assurance

Importance of Quality Assurance

The establishment of a comprehensive quality assurance program is one of the most importanttasks of a radiation oncology department. In order to determine the appropriate procedures andprocesses, various publications can be referred to that provide details on the aspects that shouldbe considered. The most comprehensive articles on this subject are the reports published by theIAEA (TRS-430 2004) and the AAPM Radiation Therapy Committee Task Group 40 (Kutcher et al,1994).

Purpose of this Document

This document is not intended as a complete guideline or working instruction. Neither does itclaim to be an all-inclusive checklist of the procedures to be completed before starting patienttreatment. It merely describes methods generally relevant to system commissioning, and providesreferences to related documents published by the international medical physics community. Theequipment, methods and tests suggested here may therefore require modification in accordancewith given standards, regulations or instructions.

Overview of Quality Assurance Procedures

Quality assurance procedures can be divided into:• Machine-related QA (see page 141)• Patient-related QA (see page 144)• Patient-specific QA (see page 146)

QUALITY ASSURANCE


10.1.1 Required Equipment

Standard Equipment Requirements

Every radiotherapy department requires certain dosimetry equipment. The following standardequipment must be available in order to facilitate the necessary commissioning procedures:• Motorized water phantom tank with control software• Various relative dose detectors (e.g. ionization chamber, diode, or diamond)• Absolute calibrated dose detector and calibrated electrometer• Solid water phantoms that are equipped with chamber drillings for absolute dose

measurements, and that also support the insertion of radiographic or radiochromatic film• Radiographic film and a film developer machine, or radiochromatic film, a calibrated film

scanner, and film analysis and dose comparison software (or 2-dimensional array with sufficientspatial resolution and control software)

• Optional: anthropomorphic phantom for verifying tissue heterogeneity

Equipment Calibration

You must ensure that the equipment in use is properly calibrated. For the purposes of comparison,several devices of a similar type should be available, e.g. two or more radiation detectors.

Test Requirements

Testing should be performed in accordance with policy guidelines determined by the hospitaldirector and by the medical physicist responsible for system operation. The tests required will varydepending on the linac and beam-collimating hardware used, and on the indications to be treatedusing this hardware. The requirements in each case may also vary in accordance with locallegislation.

Planning Software

Brainlab planning software includes tools to support the quality assurance process. The RT QAsoftware, for example, and the dose export feature, allow the dose distribution to be evaluatedand plan-measurement comparisons to be performed using third-party software.

Required Equipment


10.2 Machine-Related Quality Assurance

When is Machine-Related QA Required?

Machine-related QA must be performed whenever a part of the system is replaced or modified.You may wish to repeat machine-related QA on a regular basis (e.g. several times a year) in orderto ensure appropriate accuracy for the system as a whole.

Safety Notes

WarningMake sure that your beam data measurements are up-to-date and that dose algorithms areproperly configured and calibrated. Regularly check the configuration and calibrationusing phantom measurements.

WarningIf one or more components of the treatment delivery system have been modified,exchanged or recalibrated, a revalidation of the treatment planning system in combinationwith the treatment delivery system must be performed in accordance with your qualityassurance procedures. If components have been modified that influence the dosimetricparameters of the system, the beam data measurements must be repeated and the reviseddata entered into the system using Physics Administration.

QUALITY ASSURANCE


10.2.1 Specific Tests

Background

Machine commissioning must include testing of at least the items listed below.

Imaging Units

• Test e.g. by verifying the HU calibration of the CT scanner.• Hounsfield units converted to electron density are the basis for all dose calculation algorithms.

Imaging unit accuracy is therefore essential.

Mechanical and Kinetic Properties of the Linac System

Test the following:• Isocenter reproducibility, dependence on rotations (gantry, table, and collimator) and Winston-

Lutz testing• Accuracy and reproducibility of localization (e.g. using localizing lasers, target positioners, or

mask systems)• Accuracy and reproducibility of leaf positions and leaf movements, and leaf position

dependence on gantry position (gravity)• Collimator rotation accuracy test (star test, Rosca et al, 2006)

Beam Data and Dose Calculation

Test Type Should Include

All treatment modali-ties

• Absolute dose measurements for various field sizes covering the com-plete range of treatments

• Beam profile verification with different setups and using different detec-tors. Examples include depth dose distributions, radial/transversal pro-files, penumbra accuracy, regular and irregular irradiation fields, and rel-ative comparison of 2D and 3D dose distributions for various fields andvarious setups.

• Interleaf and intraleaf transmission (film measurement, Cosgrove et al,1999)

• Independent checks of calculated MU

Arc treatments(Grebe et al 2001)

• Gantry movements (reliability of start-stop angles, continuous move-ment, MU delivery during arcing, gantry rotation speed, dose rate de-pendence, and leaf speed dependence)

• Dynamic leaf movements• Interrupted/continued treatment

Data Transfer

• Correctness of data transfer to R&V system and linac (linac scale convention)• Correctness of data transfer to patient positioning system such as ExacTrac

Specific Tests


Light Field Test

① ②

Figure 27

This test is to ensure that machine parameters such as leaf/jaw position, collimator angle, linacangle and table angle are correctly transferred from the treatment planning system to the linac.Since RT QA is delivered with each Brainlab RT treatment planning software, the light field testonly needs to be performed in RT QA.

Step

1. Create an arc with a small couch and collimator rotation in RT QA using the Beam ModelVerification workflow (e.g. table angle 30°, collimator angle -20°, start angle 0°, stop an-gle 120° (IEC 1217)). To easily detect inverted jaw or leaf positions, define an asymmetricstart field ①. The stop field should be smaller than the start field to have the jaws aligneddirectly behind the leaf tips for the start field.

2. Transfer the treatment plan to the machine following the clinical process.

3. At the linac, use the light field and the Beam's Eye View printout ② from RT QA to verifythe correct shape of the jaw and MLC field. For start angle 0°, the light field should agreewith the printout if the printout is placed on the table (at isocenter height) and if the patienticon is correctly aligned with the table.

4. From outside the room, turn on the beam and verify correct gantry rotation and leaf move-ment.

5. Compare the light field of the stop angle with the printout again.

NOTE: Slight deviations between printout and light field may occur due to the printer calibration orbecause the MLC is calibrated to the radiation field and not the light field.

QUALITY ASSURANCE


10.3 Patient-Related Quality Assurance

When is Patient-Related QA Required?

Following machine commissioning, the quality assurance procedures performed should simulatethe complete patient treatment workflow. This should include treatments typically performed in thehospital, and also incorporate independent dose calculation at selected points in a phantom. Thedose in such cases can either be determined manually or using an alternative algorithm.

Safety Notes

WarningMeasure the absolute accuracy of the Brainlab RT Elements treatment planning software incombination with the used treatment delivery systems using phantoms. The measuredaccuracy must be taken into account when configuring plan parameters in order to ensureaccurate treatment delivery.

Patient-Related Quality Assurance


10.3.1 Recommended Procedures

Simulation of Patient Treatment

Simulation of patient treatment is recommended to be performed using an anthropomorphicphantom.• It should include all treatment steps, from CT and MR scanning through to image fusion,

treatment planning, data transfer, positioning, delivery of radiation to the phantom, dosemeasurement and comparison of the achieved results.

• Simulation should be repeated for all possible treatment modalities, indications and energylevels.

Absolute and Relative Dose Measurement

This must be performed for single arcs and the treatment plan as a whole using suitable detectorsystems, for example film with a film analysis tool, distance to agreement measures (Harms et al,1998), gamma index (Low et al, 2003), and thermoluminescence dosimetry.

QUALITY ASSURANCE


10.4 Patient-Specific Quality Assurance

When is Patient-Specific QA Required?

After machine-related QA and patient-related QA have been successfully performed, and thetreatment system is approved for patient treatment in accordance with your department’s qualitystandards, the following additional checks must be performed prior to every patient treatment.

Safety Notes

WarningEnsure proper delivery of the treatment plan to the patient. It is strongly recommended toperform a phantom verification for every treatment plan using exactly the same parametersettings that will be used for the real patient during actual treatment.

Patient-Specific Quality Assurance


10.4.1 Pre-Treatment Patient QA

Overview

Special care must be taken to ensure accurate patient positioning. To ensure correct usage ofyour stereotactic hardware and positioning system, refer to the corresponding user manual andperform the tasks described below.

Winston-Lutz Test

• Carry out a Winston-Lutz test for isocenter and laser verification at least once a day.• Ensure that all printouts are reviewed and signed by the physicist in charge.

Positioning

This applies if using Novalis Body, ExacTrac, skin markers, etc.• Position the patient based on the required isocenter using the selected system.• If applicable, verify the accuracy of the position using portal film-DRR comparison.• Check the printed beam templates for each beam using the linac light field at the specified

focus-to-film distance (for example, 1000 mm).• You should also check the correspondence of the MLC shapes with the PTV projections.

An independent check of patient positioning (for example using portal imaging, on-board imaging,or other general plausibility checks) must be performed. A second person should double checkthat positioning is correct.The entire setup accuracy depends on the laser setup. Laser verification must be performed moreoften than is usually required for standard radiotherapy. This may also be applicable for otherverification procedures.

Additional Tests

It is the responsibility of the physicist in charge to add additional tests or checks in order toguarantee the accuracy specified for the linac.

QUALITY ASSURANCE


10.4.2 General Patient QA

Recommendations

• Perform dose measurement for each patient plan, for example using a solid water phantomwith a film and absolute calibrated detectors. The results must then be compared with the dosecalculation performed using a CT scan of this phantom.

• Review all treatment parameters (e.g., mechanical properties of the linac and collimatingdevices, dosimetric properties, prescription, patient setup, and gantry-table setup) transferredto the R&V system before initial setup.

• Carry out an independent plan review.• An independent calculation of the dose at a specific point in the plan is recommended.• If applicable, verify the accuracy of patient positioning in comparison to DRRs using portal film

or EPID.

General Patient QA


11 APPENDIX 111.1 Accuracy of Dose Algorithms11.1.1 Pencil Beam and Monte Carlo

Background

The IAEA test package described in IAEA-TECDOC-1540, based on measurements by Vanselaarand Welleweerd (see the general references provided on page 155), was used to validate thedosimetric accuracy of Brainlab’s Pencil Beam and Monte Carlo implementation within BrainlabRT Elements. The test package comprises four different beam energies. The Co-60 tests have notbeen performed because Brainlab RT Elements does not support Co-machines. From theremaining three linac energies (6, 10 and 18 MV), the lowest and the highest energy were used.According to Vanselaar and Welleweerd, the 6 MV (Quality index QI = 0.676) data was measuredat an Elekta SL 15 linear accelerator, and the 18 MV (QI = 0.770) at an Elekta SL 20. Therefore,both data sets were modeled using an Elekta MLCi Standard MLC with 40 leaf pairs, each leafwith a 1 cm width.

Definitions

The deviation for a single dose point inside the open beam is calculated as(Dcalc - Dmeas) * 100% / Dmeas,and as(Dcalc - Dmeas) * 100% / Dmeas,caxfor a dose point outside the penumbra. The confidence limit is defined as|average deviation| + 1.5 * standard deviation,with the standard deviation calculated as the geometrical average of the deviations. The followingtables summarize the results for 6 and 18 MV for Pencil Beam and Monte Carlo.

IAEA Test 6 MV

Test package results for 6 MV dose calculations:

Test# Description Pencil Beam Monte Carlo

Average ConfidenceLimit Average Confidence

Limit

1a-c Square -0.1% 1% +0.1% 1.6%

2a-b Rectangular +0.2% 2.6% +0.4% 2.7%

3 Short SSD +0.0% 0.7% +0.5% 1.5%

6 Off-center plane +0.3% 1.6% +0.7% 1.9%

8a-b Lung inhomogeneity -0.2% 1.7% -0.1% 1.2%

8c Bone inhomogeneity -0.5% 1.4% +0.2% 0.9%

9 Oblique incidence +1.0% 1.6% +0.6% 1.6%

APPENDIX 1



10a-b Missing tissue +5.8% 12.6% +0.3% 1.5%

11 Asymmetrically open +2.6% 5.1% +1.2% 2.8%

IAEA Test 18 MV

Test package results for 18 MV dose calculations:


Average ConfidenceLimit Average Confidence

Limit

1a-c Square +0.0% 1.6% +0.9% 2.3%

2a-b Rectangular +0.2% 2.4% +1.1% 2.8%

3 Short SSD -0.1% 3.0% +1.6% 2.8%

6 Off-center plane +0.3% 1.9% +1.7% 2.8%

8a-b Lung inhomogeneity -0.3% 1.3% +0.2% 1.1%

8c Bone inhomogeneity +0.2% 0.8% +0.3% 1.2%

9 Oblique incidence -0.1% 1.8% +0.1% 1.7%

10a-b Missing tissue +4.5% 9.7% +0.8% 1.8%

11 Asymmetrically open +0.9% 2.1% +0.2% 1.6%

Accuracy of Dose Algorithms


11.2 Limitations of Dose Algorithms

Background

Brainlab radiotherapy treatment planning software calculates dose within clinically desirableaccuracy limits if it is used within its specifications and with parameter settings adapted to thecorresponding treatment conditions. If these conditions approach the limits of the algorithms,special care is necessary. For example, if very small MLC field sizes are used to treat patients."Very small" means that the field size is:• In the order of one or two leaf widths,• Outside the measured range of tabulated values (e.g. output/scatter factors and depth dose

tables),• Close to the Pencil Beam kernel resolution,• Close to the spatial resolution of the Monte Carlo dose calculation grid,• Close to the spatial resolution of the 3D-dose grid or• Close to the radiological correction parameters for tongue-and-groove leaf design and rounded

leaf ends.Using an inappropriate combination of these conditions, the dose can be calculated with lessaccuracy than the generally accepted standards. If this is not recognized by the user with therecommended treatment plan quality assurance, the irradiation of such a treatment plan mightlead to serious injury of the patient and/or ineffective treatment.

Extrapolation Outside the Range of Tabulated Values

The Brainlab Pencil Beam algorithm relies on tabulated values for depth dose, output factors(Scatter Factors) and off-axis profiles (Radial Factors). Arbitrary values retrieved from the tablesare interpolated accordingly. If the range of tabulated values is exceeded, certain approximationsare necessary in order to allow the display of extrapolated dose values. Naturally, the accuracy ofextrapolated values is reduced and has to be verified prior to treatment.

Resolution of Calculation Grids

Similar to other treatment planning systems, the Brainlab RT Elements treatment planningsoftware uses several calculation grid resolutions relevant for the accuracy of the dose calculation(depending on the licensed features and the TPS version):• Pencil Beam kernel resolution• Monte Carlo calculation grid resolution• 3D-dose grid resolution.

In general, the resolution of the calculation grid must be fine enough to represent the maincharacteristics of the dose distribution.Page 152 shows an exemplary dose profile of a very small radiation field, sampled with only 2 gridelements within the nominal MLC edge. As a result, the amplitude of the peak and the penumbracannot be represented with acceptable accuracy. Radiological corrections, like the Tongue-and-Groove shift (dotted green line) make this effect even stronger.Dose profile of a very small radiation field:

APPENDIX 1


Figure 28

The outer vertical line (dotted orange) represents the nominal MLC edge, while the inner verticalline (dotted green) shows the radiologic field size (position of the 50% isodose level). The red barsrepresent the profile with only two grid elements within the nominal MLC edge.To avoid unacceptable differences between the calculated and the actual dose distribution, thefield size must not be smaller than four times the grid resolution, irrespective of the dosecalculation grid type (Pencil beam kernel, Monte Carlo or 3D-dose volume). The improvement isschematically shown in the figure below. Brainlab recommends to always consider the calculationgrid resolution.Same profile as above, now sampled with four grid elements within the nominal MLC edge:

Figure 29

Effect of the Primary Collimator

WarningThe dose algorithms Pencil Beam and Monte Carlo do not explicitly model the primarycollimator (not to be confused with the primary jaws). Therefore, the dose calculationaccuracy may strongly deteriorate in the far off-axis corners of the irradiation field whichare shielded by the primary collimator (e.g. Varian linac outside an isocentric radius of220 mm). You must perform independent QA in case you intend to use those areas forirradiation of the target.

Limitations of Dose Algorithms


12 APPENDIX 212.1 Linac Energy

Nominal Linac Energy

The following table shows the most frequent photon energies. The Brainlab dose algorithmsPencil Beam and Monte Carlo are released for the Beam Quality Indices from QI = 0.61 to QI =0.80. The following table provides the related nominal linac energy.

Figure 30

Source: British Journal of Radiology - Supplement 25, “Central Axis Depth Dose Data for Use inRadiotherapy: 1996”

APPENDIX 2


Linac Energy


13 APPENDIX 313.1 Bibliography

General

• Alfonso, R., Andreo, P. et al.: A new formalism for reference dosimetry of small andnonstandard fields. Med. Phys. 35 (11); 2008; 5179-5186

• Almond, P. R., Biggs, P. J.; AAPM’s TG-51 protocol for clinical reference dosimetry of high-energy photon and electron beams. Med. Phys. 26 (9); 1999; 1847-1870

• Chang, Z., et al.: Dosimetric characteristics of Novalis Tx system with high definition multileafcollimator. Med. Phys. 35 (10); 2008; 4460-4463.

• Das, I. J., et al.: Accelerator beam data commissioning equipment and procedures: Report ofthe TG-106 of the Therapy Physics Committee of the AAPM. Med. Phys. 35 (9); 2008;4186-4215

• Das, I. J., Ding, G. X. et al.: Small fields: Nonequilibrium radiation dosimetry. Med. Phys. 35 (1);2008; 206-215

• Grebe et al.: Dynamic arc radiosurgery and radiotherapy: Commissioning and verification ofdose distributions. Int. J. Radiation Oncology Biol. Phys. 49 (5); 2001; 1451-60

• IAEA-TECDOC-1540, Specification and Acceptance Testing of Radiotherapy TreatmentPlanning Systems. IAEA; April 2006

• IAEA TRS-398; Absorbed dose determination in External Beam Radiotherapy. IAEA TechnicalReport Series No. 398; 2006;

• IAEA TRS-483; Dosimetry of Small Static Fields Used in External Beam Radiotherapy. IAEATechnical Report Series No. 483; 2017;

• IPEM Report Number 103, Small Field MV Photon Dosimetry, IPEM, December 2010• J.U. Wuerfel, Dose Measurements In Small Fields, Med Phys Int, 2013 Vol. 1 No. 1• Khan, F. M.; The Physics of Radiation Therapy, 2nd ed. Williams & Wilkins; 1993• Rice, R. K., Hansen, J. L., Svensson, G. K., Siddon, R. L.; Measurements of dose distributions

in small 6MV x-rays. Phys. Med. Biol. 32; 1987; 1087-1099• Sauer, O. A. et al.: Measurement of output factors for small photon beams. Med.Phys. 34 (6);

2007; 1983-1988• Schneider et al.; The calibration of CT Hounsfield Units for radiation treatment planning. (1996)

Phys. Med. Biol. 41 pp 111-124.• Sterling et.al., Automation of radiation treatment planning; Brit. J. Radiol. 37, 544 (1964).• Vanselaar, J. and Welleweerd, H., Application of a test package in an intercomparison of the

photon dose calculation performance of treatment planning systems used in a clinical setting.Radiotherapy & Oncology 60; 2001; 203-213

• Winkler P. et al.: Dose-response characteristics of an amorphous silicon EPID. Med. Phys. 32(10); 2005; 3095-3105

• Yin et al.: Dosimetric characteristics of the Novalis shaped beam surgery unit. Med. Phys. 29(8); 2002; 1729-38

Quality Assurance Bibliography

• Cosgrove et al.: Commissioning of a multi-leaf collimator and planning system for stereotacticradiosurgery. Radiother. Oncol. 50; 1999; 325-36

APPENDIX 3


• Grebe et al.: Dynamic arc radiosurgery and radiotherapy: Commissioning and verification ofdose distributions. Int. J. Radiation Oncology Biol. Phys. 49 (5); 2001; 1451-60

• Harms, W. B., et al.: A software tool for the quantitative evaluation of 3D dose calculationalgorithms. Med. Phys. 25 (10); 1998; 1830-1836

• IAEA TRS-430; Commissioning and Quality Assurance of Computerized Planning Systems forRadiation Treatment of Cancer. IAEA Technical Report Series No. 430; 2004

• Kutcher, G. J., et al.: Comprehensive QA for Radiation Oncology: Report of AAPM RadiationTherapy Committee Task Group 40. Med. Phys. 21 (4); 1994; 581-618

• LoSasso, T. et al.: Comprehensive quality assurance for the delivery of intensity modulatedradiotherapy with a multileaf collimator used in the dynamic mode. Med. Phys. 28 (11); 2001;2209-2219

• LoSasso, T. et al.: Physical and dosimetric aspects of a multileaf collimation system used in thedynamic mode for implementing intensity modulated radiotherapy. Med. Phys. 25 (10); 1998;1919-1927

• Low, D. A. et al.: Evaluation of the gamma dose distribution comparison method. Med. Phys. 30(9); 2003; 2455-2464

• Rosca, F., et al.: An MLC-based Linac QA procedure for the characterization of radiationisocenter and room lasers' position. Med. Phys. 33 (6); 2006; 1780-1787

• Zygmanski, P. et al.: Dependence of fluence errors in dynamic IMRT on leaf-positional errorsvarying with time and leaf number. Med. Phys. 30 (10); 2003; 2736-2749

Quality Assurance - Further Reading

• AAPM Reports (http://www.aapm.org/pubs/reports/)• Agazaryan et al.: Patient specific quality assurance for the delivery of intensity modulated

radiotherapy. J. Appl. Clin. Med. Phys. 4 (1); 2003; 40-50• Agazaryan et al.: A methodology for verification of radiotherapy dose calculation. J. Neurosurg.

(Suppl. 3) 101; 2004; 356-61• Clark et al.: Penumbra evaluation of the Varian Millenium and the Brainlab m3 multileaf

collimators. Int. J. Radiation Oncology Biol. Phys. 66 (4); 2006; S71-S75• Leavitt et al.: Comparison of interpolated vs. calculated micro-multileaf settings in dynamic

conformal arc treatment. Med. Dosim. 25 (1); 2000; 17-21• Li et al.: A new approach in dose measurement and error analysis for narrow photon beams

(beamlets) shaped by different multileaf collimators using a small detector. Med. Phys. 31 (7);2004; 2020-32

• Linthout et al.: Evaluation of dose calculation algorithms for dynamic arc treatments of headand neck tumors. Radiother. Oncol. 64; 2002; 85-95

• Linthout et al.: A simple verification of monitor unit calculation for intensity modulated beamsusing dynamic mini-multileaf collimation. Radiother. Oncol. 71; 2004; 235-41

• Solberg et al.: Conformal radiosurgery using a dedicated Linac and micro multileaf collimator.In: Kondziolka (ed) Radiosurgery; Basel; Karger; vol. 3; 2000; 53-63

• Verellen et al.: Assessment of the uncertainties in dose delivery of a commercial system forLinac-based radiosurgery. Int J Radiation Oncology Biol Phys 44 (2); 1999; 421-33

• Wong: Quality assurance devices for dynamic conformal radiotherapy. J. Appl. Clin. Med. Phys.5 (1); 2004; 1-8

• Xia et al.: Physical characteristics of a miniature multileaf collimator. Med. Phys. 26 (1); 1999;65-70

Pencil Beam Bibliography

• Mohan R, Chui C, Lidofsky L; Energy and angular distributions of photons from medical linearaccelerators. (1985) Med. Phys. 12 pp 592 - 597.

• Mohan R, Chui C, Lidofsky L; Differential pencil beam dose computation model for photons.(1986) Med. Phys. 13 pp 64 - 73.

• Mohan R, Chui C; Use of fast fourier transforms in calculating dose distributions for irregularlyshaped fields for three-dimensional treatment planning. (1987) Med. Phys. 14 pp 70 - 77.

Bibliography


Monte Carlo Bibliography

• AAPM Task Group Report No 105: Issues associated with clinical implementation of MonteCarlo-based external beam treatment planning, Medical Physics 34 (2007) 4818-4853.

• Berger M J, Hubbell J H: XCOM: Photon cross sections on a personal computer, TechnicalReport NBSIR 87-3597 (1987) National Institute of Standards and Technology, Gaithersburg,MD.

• Berger M J: ESTAR, PSTAR, and ASTAR: Computer programs for calculating stopping-powerand range tables for electrons, protons, and helium ions, Technical Report NBSIR 4999 (1993)National Institute of Standards and Technology, Gaithersburg MD.

• Dobler B, Walter C, Knopf A, Fabri D, Loeschel R, Polednik M, Schneider F, Wenz F, Lohr F:Optimization of extracranial stereotactic radiation therapy of small lung lesions using accuratedose calculation algorithms, Radiation Oncology 1 (2006) 45.

• Fippel M: Fast Monte Carlo dose calculation for photon beams based on the VMC electronalgorithm, Medical Physics 26 (1999) 1466-1475.

• Fippel M: Efficient particle transport simulation through beam modulating devices for MonteCarlo treatment planning, Medical Physics 31 (2004) 1235-1242.

• Fippel M, Haryanto F, Dohm O, Nüsslin F, Kriesen S: A virtual photon energy fluence model forMonte Carlo dose calculation, Medical Physics 30 (2003) 301-311.

• Fippel M, Kawrakow I, Friedrich K: Electron beam dose calculations with the VMC algorithmand the verification data of the NCI working group, Physics in Medicine and Biology 42 (1997)501-520.

• Fippel M, Laub W, Huber B, Nüsslin F: Experimental investigation of a fast Monte Carlo photonbeam dose calculation algorithm, Physics in Medicine and Biology 44 (1999) 3039-3054.

• ICRU Report No 46: Photon, Electron, Proton and Neutron Interaction Data for Body Tissues,International Commission on Radiation Units and Measurements (1992).

• International Electrotechnical Commission: Radiotherapy equipment, coordinates, movementand scales, IEC 1217 (1996).

• Kawrakow I: Accurate condensed history Monte Carlo simulation of electron transport. I.EGSnrc, the new EGS4 version, Medical Physics 27 (2000) 485-498.

• Kawrakow I, Fippel M: Investigation of variance reduction techniques for Monte Carlo photondose calculation using XVMC, Physics in Medicine and Biology 45 (2000) 2163-2183.

• Kawrakow I, Fippel M, Friedrich K: 3D Electron Dose Calculation using a Voxel based MonteCarlo Algorithm (VMC), Medical Physics 23 (1996) 445-457.

• Krieger T, Sauer O A: Monte Carlo- versus pencil-beam-/collapsed-cone-dose calculation in aheterogeneous multi-layer phantom, Physics in Medicine and Biology 50 (2005) 859-868.

• Press W H, Flannery B P, Teukolsky S A, Vetterling W T: Numerical Recipes in C: The Art ofScientific Computing, Second Edition, Cambridge University Press (1992).

• Reynaert N, van der Marck S C, Schaart D R, Van der Zee W, Van Vliet-Vroegindeweij C,Tomsej M, Jansen J, Heijmen B, Coghe M, De Wagter C: Monte Carlo treatment planning forphoton and electron beams, Radiation Physics and Chemistry 76 (2007) 643-686.

• Rogers D W O, Faddegon B A, Ding G X, Ma C M, We J, Mackie T R: BEAM: A Monte Carlocode to simulate radiotherapy treatment units, Medical Physics 22 (1995) 503-524.

• Vanderstraeten B, Chin P W, Fix M, Leal M, Mora G, Reynaert N, Seco J, Soukup M, Spezi E,De Neve W, Thierens H: Conversion of CT numbers into tissue parameters for Monte Carlodose calculations: a multi-centre study, Physics in Medicine and Biology 52 (2007) 539-562.

Dynamic Delivery Bibliography

• Bedford JL and Warrington AP, Commissioning of Volumetric Modulated Arc Therapy (VMAT),Int J Rad Oncol Biol Phys 73(2), 537 (2009)

• Clifton Ling C et al., Commissioning and Quality Assurance of RapidArc Radiotherapy DeliverySystem, Int J Rad Oncol Biol Phys 72(2), 575 (2008)

• Hernandez V et al., Determination of the optimal tolerance for MLC positioning in slidingwindow and VMAT techniques, Med Phys 42, 1911 (2015)

• Kielar K et al., Verification of dosimetric accuracy on the TrueBeam STx: Rounded leaf effect ofthe high definition MLC, Med Phys 39, 6360 (2012)

APPENDIX 3


• Masi L et al., Quality assurance of volumetric modulated arc therapy: Evaluation andcomparison of different dosimetric systems, Med Phys 38, 612 (2011)

• Van Esch A et al., Implementing RapidArc into clinical routine: A comprehensive program frommachine QA to TPS validation and patient QA, Med Phys 38, 5146 (2011)

• Wang Q, Dai J and Zhang K, A novel method for routine quality assurance of volumetric-modulated arc therapy, Med Phys 40, 101712 (2013)

• Yao W and Farr JB, Determining the optimal dosimetric leaf gap setting for rounded leaf-endmultileaf collimator systems by simple test fields, J App Clin Med Phys 16(4) (2015)

Bibliography


INDEXA

absolute linac calibrationpencil beam............................................................................ 47

adaptive grid...............................................................................38

Bbackground leakage

pencil beam............................................................................ 50beam data for brainlab m3

monte carlo...........................................................................122pencil beam............................................................................ 65

beam data for elekta agilitymonte carlo........................................................................... 117pencil beam............................................................................ 69

beam data for MHI MLC 60monte carlo...........................................................................120pencil beam............................................................................ 73

beam data for novalispencil beam............................................................................ 76

beam data for varian 120 (SRS flattening filter)monte carlo...........................................................................128pencil beam............................................................................ 88

beam data for varian 120 (standard irradiation and flattening filterfree mode)monte carlo...........................................................................130

beam data for varian 120 (standard irradiation mode andflattening filter free mode)pencil beam............................................................................ 92

beam data for varian HD120 (SRS flattening filter)monte carlo...........................................................................124pencil beam............................................................................ 80

beam data for varian HD120 (standard irradiation and flatteningfilter free mode)monte carlo...........................................................................126

beam data for varian HD120 (standard irradiation mode andflattening filter free mode)pencil beam............................................................................ 84

bibliography.............................................................................. 155Brainlab imaging couch top........................................................ 17

Ccarbon fiber

dose build-up.......................................................................... 17CE certification............................................................................. 8coordinate systems

monte carlo........................................................................... 112

Ddata correction

monte carlo........................................................................... 113depth dose profile

pencil beam............................................................................ 51dose algorithms

limitations..............................................................................151dose build-up..............................................................................17dose detectors..........................................................................140

recommended equipment, monte carlo.................................111recommended equipment, pencil beam................................. 45

dynamic conformal arc............................................................... 39dynamic leaf shift

pencil beam............................................................................ 62

Ffield measurement......................................................................21field setup................................................................................... 17film dosimetry measurement

pencil beam............................................................................ 61

Hhigh-resolution detector..............................................................60

IIAEA test package....................................................................149imaging couch top...................................................................... 17indications for use...................................................................... 10

Jjaw measurement

pencil beam............................................................................ 50

Lleaf positions............................................................................ 102linac energy.............................................................................. 153

Mmachine-related quality assurance...........................................141measurement requirements

pencil beam............................................................................ 46measurements in air

monte carlo........................................................................... 114measurements in water

monte carlo........................................................................... 116monte carlo

dose computation engine..................................................... 103modeling of the collimating system.......................................101virtual energy fluence model...................................................99x-ray voxel.............................................................................. 98

monte carlo parameters........................................................... 105

Nnominal linac output measurement

pencil beam............................................................................ 47

Ooutput factors

pencil beam............................................................................ 53

Ppatient-related quality assurance............................................. 144

INDEX


patient-specific quality assurance............................................ 146PDD and TPR measurement

pencil beam............................................................................ 51pencil beam

adaptive grid calculation.................................................... 31,38fast fourier transformation.......................................................31fluency distribution..................................................................31limitations...........................................................................40,41path length correction........................................................ 31,37

pencil beam theoryconvolution............................................................................. 33differential............................................................................... 31fluence matrix......................................................................... 33idealized dose distribution...................................................... 33introduction............................................................................. 31kernel......................................................................................32monitor unit calibration........................................................... 36monoenergetic........................................................................31nominal linac output................................................................34radial factors........................................................................... 35source function correction...................................................... 33source-isocenter distance.......................................................34source-surface distance......................................................... 34tissue phantom ratio............................................................... 36total dose................................................................................ 34total scatter factor................................................................... 36

phantoms..................................................................................140physics administration

pencil beam............................................................................ 25raw data..................................................................................24

Qquality assurance..................................................................... 139

equipment calibration........................................................... 140equipment requirements.......................................................140machine-related....................................................................141patient-related.......................................................................144patient-specific......................................................................146phantoms..............................................................................140

Rradial profiles

pencil beam............................................................................ 55radiation fields............................................................................ 21radiochromatic film................................................................... 140radiographic film....................................................................... 140ray tracing...................................................................................37recommended equipment

monte carlo........................................................................... 111recommended measurement equipment

pencil beam............................................................................ 45

Sscatter factor measurement

pencil beam............................................................................ 53skin dose build-up...................................................................... 17support numbers.......................................................................... 7symbols........................................................................................ 9

Ttissue maximum ratio................................................................. 37tissue phantom ratio................................................................... 36trademarks

Brainlab.................................................................................... 8third-Party................................................................................. 8

transversal profilepencil beam............................................................................ 58

Uuser guides................................................................................. 11

Wwater phantom............................................................................60

INDEX


brainlab.com

Art-No. 60919-42EN

*60919-42EN*

Documents

PHYSICS BRAINLAB · Technical Reference Guide Rev. 2.2 Brainlab Physics 9 1.4 Using the System Purpose and Audience • This guide provides a background and reference for the medical