1

Quality Assurance of Radiation Treatment Planning Systems:A Significant Challenge

Jake Van DykLondon Regional Cancer ProgramUniversity of Western Ontario

University of Western Ontario

!?

7/20/2005125 July 20051

Introduction

• Technological revolution in radiation oncology• Enhanced use of imaging

• Computer-controlled dose delivery

• Tighter margins• Higher doses

• Dynamic delivery

• Central to this is the treatment planning system (TPS)

7/20/2005225 July 20052

Introduction

• Modern TPS• Increased use of patient images

• Possibly from various imaging modalities

• Enhanced 3-D displays• More sophisticated dose calculation algorithms

• More complex treatment plan evaluation tools

• Generation of images used for treatment verification

• IMRT• Step and shoot

• Dynamic delivery

7/20/2005325 July 20053

Educational Objectives

1. To demonstrate the importance of QA of TPSs by reviewing significant treatment errors associated with their use.

2. To review the major functionality of a modern TPS.3. To highlight and summarize various reports that have made

recommendations regarding commissioning and QA of TPSs.4. To discuss accuracy requirements and criteria of acceptability of

the modern TPS.5. To summarize acceptance testing procedures as proposed by the

IAEA for a modern TPS.

6. To provide an overview of commissioning a modern RTPS.7. To provide an overview of the QA associated with a modern TPS.

2

7/20/2005425 July 20054

Recent References

Med Phys 25:1773-829,1998

7/20/2005525 July 20055

Recent References

1999

7/20/2005625 July 20056

Recent References

• For manufacturers

International ElectrotechnicalCommission (IEC), 2000

7/20/2005725 July 20057

Recent References

Available in pdf format from:

http://www-pub.iaea.org/MTCD/publications/PDF/TRS430_web.pdf

2004

3

7/20/2005825 July 20058

IAEA TRS 430 Contents

1. Introduction2. Clinical treatment planning process3. Description of radiation treatment planning systems4. Algorithms used in radiation treatment planning5. Quality assessment6. Quality assurance management7. Purchase process8. Acceptance testing9. Commissioning10. Periodic quality assurance11. Patient-specific quality assurance12. Summary

7/20/2005925 July 20059

Recent References

• ESTRO 2004

Available from ESTRO website:http://www.estroweb.org/estro/index.cfm

7/20/20051025 July 200510

VeryVery Recent Reference

• Includes chapters on:• Imaging for treatment planning

• Inverse planning

• Monte Carlo for treatment planning• IMRT

• Radiobiological modeling

• Breathing control in radiation therapy• Prostate brachytherapy

25 July 2005

7/20/20051125 July 200511

Future Reference

4

7/20/20051225 July 200512

Future Reference

2006?

7/20/20051325 July 200513

QA in Radiation Therapy (RT)• Two considerations in RT

Avoidance of treatment errors

0 10 20 30 40 50 60 70 80 900

20

40

60

80

100

Normal Tissue Tumor

Dose (Gy)

Rel

ativ

e R

espo

nse

[%]

TumorNormal tissue

Need for accuracy in RT process

7/20/20051425 July 200514

Need for Accuracy in Dose Calculations

Uncertainty Type UncertaintyRange (%)

A Absorbed dose to referencepoint in water phantom

2.5

B Determination of relative dose(Measurement away fromreference point)

2.5

C Relative dose calculations 2.5D Patient irradiation 2.5E Overall 5.0

• General accuracy desired in dose delivered to patient: 5%

7/20/20051525 July 200515

ICRU Goal in Dose Calculation and Spatial Accuracy

• Relative dose accuracy in uniform dose region: 2%

• Spatial accuracy in high dose gradient: 2 mm

Off Axis profile

0

10

20

30

40

50

60

70

80

90

100

110

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

Off axis distance (cm)

Rea

lati

ve D

ose

Measured Computed

2%

4 mm

Real DataICRU 42, 1987

5

7/20/20051625 July 200516

Avoidance of Treatment Errors

• Error• “The failure of planned action to be completed as intended (i.e., error of execution) or the use of a wrong plan to achieve an aim (i.e., error of planning).”

Institute of Medicine. To Err is Human: Building a Safer Health System, 2000.

7/20/20051725 July 200517

Avoidance of Errors in RT

IAEA 2000 ICRP 2000

7/20/20051825 July 200518

IAEA: Categories of Errors

Categories Number oferrors

Radiation measurement systems 5External beam: Machine commissioning & calibration 15External beam therapy: Treatment planning, patient setup and treatment 26Decommissioning of teletherapy equipment 2Mechanical and electrical malfunctions 4Brachytherapy: Low dose rate sources and applicators 29Brachytherapy: High dose rate 3Unsealed sources 8

92

7/20/20051925 July 200519

IAEA: Lessons… Examples

Description CommentsInconsistent/incorrect data set Lack of proper commissioning/verificationInsufficient understanding of algorithm Lack of understanding on use of wedge

factorsIncorrect calculation of treatment times Lack of independent checkIncorrect distance correction Lack of understanding/training

Lack of independent checkMisunderstanding of complex treatment plan - verbal communication

Lack clear documentationIneffective communication

Incorrect positioning of beams on patient

Poor implementation of instructions

Wrong source strength Insufficient tranining/understandingNo independent check

Wrong isotope No independent checkError in removal time No independent check

6

7/20/20052025 July 200520

IAEA

7/20/20052125 July 200521

Discovery of Problem

• Nov 2000, patients with prolonged diarrhoea

• Physicists reviewed plans - no anomaly• Double checking of plans was used

• No independent, manual time calc check

• Dec 2000, similar symptoms in other patients

• Feb 2001, physicists initiated search for cause of such effects

• March 2001, physicists discovered problem with the calculation of treatment times…

7/20/20052225 July 200522

Treatment Planning Problem

• “Shortcut” method of block entry even when 4 blocks

Two approaches

Time incorrect Time correct7/20/200523

25 July 200523

Identification of Problem

Open beam

50%

30%

70%

90%

Blocks entered together with two CW contours

7

7/20/20052425 July 200524

Problem Summary

• “These factors, together with an apparent omission of manual checking of computer calculations, resulted in the patients concerned being exposed at radiation levels that were set too high.”

7/20/20052525 July 200525

Clinical Summary - March 2002

• 28 patients treated with incorrect doses• 17 have since died• 13 had rectal complications

7/20/20052625 July 200526

Physicists Jailed

November 2004• Two physicists

• Condemned to 4 years in prison• Barred to practice their profession for 7 years

• Third physicist is acquitted

7/20/20052725 July 200527

Errors in RT: Contributing Factors

• Insufficient education

• Lack of procedures/protocols as part of comprehensive QA program

• Lack of supervision of compliance with QA program

• Lack of training for “unusual” situations

• Lack of a “safety culture”

8

7/20/20052825 July 200528

Hardware• CPU• High resolution graphics• Mass storage (hard disc)• Floppy disk/CD ROM• Keyboard & mouse• High resolution monitor• Digitizer• Laser/color printer• Backup storage facility• Network connections

Components of 3-D TPS

7/20/20052925 July 200529

Software• Input routines

• Anatomy modeling• Beam geometry (virtual simulation)

• Dose calculations• Dose volume histograms/evaluation tools

• Digitally reconstructed radiographs

• Output [hardcopies, network, web connection (RTOG)]

Components of 3-D TPS

7/20/20053025 July 200530

Components of 3-D TPSSoftware utilities for…• Treatment unit data/dose data• Brachytherapy isotope data• Output of treatment unit/source data• Organization of patient data• Contouring software – targets, organs• Video display for interactive beam placement, shaping, sizing• Dose calculation initialization – grid, type of dose algorithm• Dose calculation software• Isodose display software incl. normalization, beam weights• Hardcopy software• Archiving software• Backup software

7/20/20053125 July 200531

Algorithm Classification

• All algorithms have two major attributes:

A. Scatter Integral• Release and Spread of Energy

• Primary and scatter

• Level of summation or integration

B. Patient Anatomy• Use of Patient Anatomical Data

• Imaging geometry and density data

• Assumptions of symmetry

9

7/20/20053225 July 200532

A. Scatter IntegralSuperposition Principle

Point Kernel

Slab Kernel

Beam Kernel

Pencil Kernel

7/20/20053325 July 200533

B. Use of Anatomy Data

• Patient’s Anatomy• As imaged by CT, MR, PET, etc

• Geometry and density • As sensed by algorithm

• Symmetry assumptions

• 1-D, 2-D, 2.5-D, or 3-D matrix

7/20/20053425 July 200534

If you are buying...

• Does the algorithm “feel” the anatomy voxels in 1-D? 2-D? 3-D?

• Is the algorithm• 0-D,1-D, 2-D or 3-D in its scatter

integration?

• Does the algorithm handle electron transport?

• Does it use a point kernel or a pencil beam kernel?

• How is the primary TERMA and kernel changed with density? Atomic number?

Of course we use

a 3-D convolution!

7/20/20053525 July 200535

http://www.opctonline.net/- Interesting background information

10

7/20/20053625 July 200536

http://www.opctonline.net/

7/20/20053725 July 200537

7/20/20053825 July 200538

Components of QA Program

• Program & system documentation• User training

• Sources of uncertainties• Suggested tolerances

• Initial system checks (commissioning)• QC - repeated system checks

• QC - “manual” checks (patient specific)• QC - in vivo dosimetry

• QA - administration

7/20/20053925 July 200539

Commissioning and QA of TPS

User Manufacturer User

Input Output

Data(Radiation,

Patient)

RadiationTherapyPlanningSystem

DoseDistributions

Compare to expected results

Commissioning: Initial testsQuality control: Reproducibility

11

7/20/20054025 July 200540

Regions of Different Dose Calculation Accuracy

AAPM TG537/20/200541

25 July 200541

Sample Criteria of Acceptability

IAEA TRS 430

7/20/20054225 July 200542

Acceptance Testing

• What happens in reality!• Catalogue delivered components

• Hardware• Software

• Test components for functionality• Sign acceptance document

• What should happen!• Based on IAEA consultants meeting 14-18 March 2005• Manufacture to perform series of type tests (e.g., TG23/Report 55)• Type test results should be documented and made available to user• Site (acceptance) tests should be a subset of type tests performed at

the time of TPS installation• Results compared to results of type tests

7/20/20054325 July 200543

Type Tests

• Elekta

• 6, 10, 18 MV

Venselaar & Welleweerd Radioth Oncol 60: 203-213, 2001.

12

7/20/20054425 July 200544

Commissioning

• Prepare system for clinical use• Provides experience/training for users

• Enter appropriate measured data• %DD, TAR, TPR, beam profiles, wedge profiles, attenuation data,

output factors, etc

• Perform series of commissioning tests• Tests algorithms

• Provides capabilities & limitations

• Assess results to see if they comply with specifications• Provides documentation of system performance

• Results of commissioning tests used later for QC tests

7/20/20054525 July 200545

Commissioning

• IAEA TRS-430 provides sample tests• System set-up/machine configuration

• Patient anatomical representation

• External beam commissioning• Brachytherapy commissioning

• Plan evaluation tools

• Plan output and data transfer• Overall clinical tests

7/20/20054625 July 200546

IAEA TRS 4307/20/200547

25 July 200547

Phantoms

• IMRT & 3-D QA

Med-Tec

13

7/20/20054825 July 200548

QA of the Non-Dosimetric Components of a TPS: QUASAR®

DisclosureDisclosure• The QUASAR® QA phantoms about to be described are

commercial products• Invented by J Van Dyk, T Craig, D Brochu, A McNiven, T Kron• Marketed by Modus Medical Devices, London, Ontario

• www.modusmed.com

• References: • A Quality Assurance Phantom for Three-Dimensional Radiation

Treatment Planning. Craig, T.C., Brochu, D., Van Dyk, J. Int. J. Radiat. Oncol. Biol. Phys., 44, 955-966, 1999.

• A Multileaf Collimator Phantom for the Quality Assurance of Radiation Therapy Planning Systems and CT Simulators. McNiven, A., Kron, T., Van Dyk, J. Int. J. Radiat. Oncol. Phys. Biol. 60, 994-1001, 2004.

7/20/20054925 July 200549

Non-Dosimetric Issues

• Image acquisition and transfer• Beam display

• CT image reconstructions• Multiplanar CT image reconstructions

• Digitally reconstructed radiographs

• Anatomical volumes• 3-D display

• Automatic tools - autocontouring, automargin, etc

• Dose volume histograms

• CT numbers to electron density conversion

7/20/20055025 July 200550

Phantom Schematic

7/20/20055125 July 200551

Commercial Version of QA Phantom

Modus Modus Medical Devices, London, Ontario

14

7/20/20055225 July 200552

7/20/20055325 July 200553

Reconstructed Image Verification

Oblique CT reconstruction Digitally reconstructed radiograph

7/20/20055425 July 200554

Multi-Institution Evaluation

• Phantom used to evaluate 3 TPSs and 1 CT simulator• Picker ACQSIM• Varian CADPlan

• ADAC Pinnacle

• Theratronics TheraplanPlus

7/20/20055525 July 200555

2cm

5cm

15

0

1

2

A B C D

110120130140150

A B C D

0

2 0

4 0

A B C D

02 04 06 0

A B C D

P o l y s t y r e n e C y l i n d e rP o l y s t y r e n e C u b e

L u c i t e C u b eA i r W e d g e

M a x i m u m v a r i a t i o n + 4 2 % t o -4 4 %

V o l u m e M e a s u r e m e n t s ( c m3)

7 / 2 0 / 2 0 0 55 72 5 J u l y 2 0 0 55 7

D o s e V o l u m e H i s t o g r a m s

0 2 04 06 08 01 0 01 2 0

0

2 0

4 0

6 0

8 0

1 0 0

C a l c u l a t e dM e a s u r e d

Vo

lu

me

(

%)

D o s e ( % )

7 / 2 0 / 2 0 0 55 82 5 J u l y 2 0 0 55 8

0 . 00 . 51 . 01 . 52 . 02 . 5- 1 0 0 0

- 5 0 0

0

5 0 0

1 0 0 0

1 5 0 0

2 0 0 0 S y s t e m A S y s t e m B S y s t e m C B a t t i s t a a n d B r o n s k i l l

CT

N

um

be

r

R e l a t i v e E l e c t r o n D e n s i t y

C T N u m b e r t o E l e c t r o n D e n s i t y C o n v e r s i o n

7 / 2 0 / 2 0 0 55 92 5 J u l y 2 0 0 55 9

M L C V e r s i o n

A c r y l i c

A i r

• V a r i a n 5 2 , 8 0 a n d 1 2 0 L e a f

• E l e k t a• R a d i o n i c sm i c r o-M L C

• S i e m e n s• V a r i a n 1 2 0 L e a f• B r a i n l a bm i c r o-M L C

16

7/20/20056025 July 200560

MLC Phantom AAPM TG 66, 2003

7/20/20056125 July 200561

Multi-Observer Test

• Errors ≥≥≥≥ 2 mm, identified 100% of the time

• 1 mm errors were identified 80% of the time

Does leaf end align with phantom geometry (air/acrylic interface)?

7/20/20056225 July 200562

Non-Dosimetric Components

• Non-dosimetric components require QC• Phantom is a unique tool for QC

• 3-D TPS

• CT-simulator

• Allows assessment of errors, limitations and uncertainties of 3-D TPS

• Several problems discovered in various commercial 3-D TPS software

7/20/20056325 July 200563

Quality ControlPS = Patient specific, W = Weekly, M =Monthly, Q = Quarterly, A = Annually,U = After software or hardware update

IAEA TRS-430 Table 611 Sonic digitizer, 2 Electromagnetic digitizer

17

7/20/20056425 July 200564

• One “qualified medical physicist” responsible• Documentation of QA process

• Record results• Clear channels of communication re:

• Software changes on TPS

• New/altered data files• CT imager software/hardware changes

• Machine output changes

QA Administration

7/20/20056525 July 200565

• Formal QC program includes:• User training• Well-defined (re)commissioning tests• Well-defined repeatability checks• Appropriate actions as needed• Documentation of results• Patient specific QC

• Process QA• Incident/error rate• Number of replans• Timeliness• Physician satisfaction

Summary

7/20/20056625 July 200566

Take Home Message

• TPS consists many components• Hardware

• Software

• Acceptance and commissioning will take time and effort• Usually considers dose calculation issues• Should also includes imaging issues

• QA of non-dose components is also important

7/20/20056725 July 200567

Key Issues

CommunicationVerification

DocumentationEducation

18

7/20/20056825 July 200568

Summary

• Quality assurance of radiation treatment planning systems is a significant challenge• but not insurmountable!