Introduction
•
Enhanced use of imaging
Computer-controlled dose delivery
• Tighter margins
• Higher doses
• Dynamic delivery
•
Quality Assurance of Radiation
Treatment Planning Systems:
A Significant Challenge ?
!
•
Jake Van Dyk
London Regional Cancer Program
University of Western Ontario
University of
Western Ontario
•
1
Educational Objectives
Modern TPS
•
Increased use of patient images
•
Enhanced 3-D displays
More sophisticated dose calculation algorithms
More complex treatment plan evaluation tools
Generation of images used for treatment verification
IMRT
•
•
•
•
•
•
•
1.
2.
Possibly from various imaging modalities
3.
4.
5.
Step and shoot
Dynamic delivery
6.
7.
2
2
7/20/2005
25 July 2005
Central to this is the treatment planning system (TPS)
1
7/20/2005
25 July 2005
Introduction
•
Technological revolution in radiation oncology
3
To demonstrate the importance of QA of TPSs by reviewing
significant treatment errors associated with their use.
To review the major functionality of a modern TPS.
To highlight and summarize various reports that have made
recommendations regarding commissioning and QA of TPSs.
To discuss accuracy requirements and criteria of acceptability of
the modern TPS.
To summarize acceptance testing procedures as proposed by the
IAEA for a modern TPS.
To provide an overview of commissioning a modern RTPS.
To provide an overview of the QA associated with a modern TPS.
3
7/20/2005
25 July 2005
1
Recent References
Recent References
1999
Med Phys 25:1773-829,1998
4
4
7/20/2005
25 July 2005
5
Recent References
•
Recent References
2004
For manufacturers
International Electrotechnical
Commission (IEC), 2000
6
5
7/20/2005
25 July 2005
6
7/20/2005
25 July 2005
Available in pdf format from:
http://www-pub.iaea.org/MTCD/publications/PDF/TRS430_web.pdf
7
7
7/20/2005
25 July 2005
2
IAEA TRS 430 Contents
Recent References
•
Introduction
2. Clinical treatment planning process
3. Description of radiation treatment planning systems
4. Algorithms used in radiation treatment planning
5. Quality assessment
6. Quality assurance management
7. Purchase process
8. Acceptance testing
9. Commissioning
10. Periodic quality assurance
11. Patient-specific quality assurance
12. Summary
1.
8
8
7/20/2005
25 July 2005
Available from ESTRO website:
http://www.estroweb.org/estro/index.cfm
9
Very Recent Reference
•
ESTRO 2004
9
7/20/2005
25 July 2005
Future 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
10
10
7/20/2005
25 July 2005
11
11
7/20/2005
25 July 2005
3
Future Reference
QA in Radiation Therapy (RT)
•
Two considerations in RT
Need for accuracy in
RT process
Avoidance of
treatment errors
100
Relative Response [%]
80
2006?
Normal
tissue
Normal Tissue
60
Tumor
Tumor
40
20
0
0
10
20
30
40
50
60
70
80
90
Dose (Gy)
12
7/20/2005
25 July 2005
13
7/20/2005
25 July 2005
13
Need for Accuracy in Dose
Calculations
•
General accuracy desired in dose delivered to patient: 5%
Uncertainty Type
A Absorbed dose to reference
point in water phantom
B Determination of relative dose
(Measurement away from
reference point)
C Relative dose calculations
D Patient irradiation
E Overall
14
ICRU Goal in Dose Calculation
and Spatial Accuracy
14
7/20/2005
25 July 2005
ICRU 42, 1987
Relative dose
accuracy in
uniform dose
region: 2%
• Spatial
accuracy in
high dose
gradient: 2 mm
•
Uncertainty
Range (%)
2.5
2.5
2.5
2.5
5.0
Off Axis profile
110
Real Data
100
2%
90
80
Realative Dose
12
70
4 mm
60
50
40
30
20
Measured
10
Computed
0
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)
15
15
7/20/2005
25 July 2005
4
Avoidance of Treatment Errors
•
Avoidance of Errors in RT
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.
IAEA 2000
16
16
7/20/2005
25 July 2005
17
IAEA: Categories of Errors
Categories
Radiation measurement systems
External beam:
Machine commissioning & calibration
External beam therapy:
Treatment planning, patient setup and treatment
Decommissioning of teletherapy equipment
Mechanical and electrical malfunctions
Brachytherapy:
Low dose rate sources and applicators
Brachytherapy: High dose rate
Unsealed sources
18
18
7/20/2005
25 July 2005
ICRP 2000
17
7/20/2005
25 July 2005
IAEA: Lessons… Examples
Number of
errors
5
15
26
2
4
29
3
8
92
Description
Inconsistent/incorrect data set
Insufficient understanding of algorithm
Incorrect calculation of treatment times
Incorrect distance correction
Misunderstanding of complex treatment
plan - verbal communication
Incorrect positioning of beams on
patient
Wrong source strength
Wrong isotope
Error in removal time
19
Comments
Lack of proper commissioning/verification
Lack of understanding on use of wedge
factors
Lack of independent check
Lack of understanding/training
Lack of independent check
Lack clear documentation
Ineffective communication
Poor implementation of instructions
Insufficient tranining/understanding
No independent check
No independent check
No independent check
19
7/20/2005
25 July 2005
5
Discovery of Problem
IAEA
•
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
…
20
20
7/20/2005
25 July 2005
21
Identification of Problem
Treatment Planning Problem
•
21
7/20/2005
25 July 2005
“Shortcut” method of block entry even when 4 blocks
Two approaches
90%
70%
Open beam
50%
30%
Time incorrect
22
22
7/20/2005
25 July 2005
Blocks
entered
together with
two CW
contours
Time correct
23
23
7/20/2005
25 July 2005
6
Problem Summary
•
24
Clinical Summary - March 2002
“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.”
24
7/20/2005
25 July 2005
•
•
•
25
Physicists Jailed
26
7/20/2005
25 July 2005
17 have since died
13 had rectal complications
25
7/20/2005
25 July 2005
Errors in RT: Contributing Factors
November 2004
• Two physicists
• Condemned to 4 years in prison
• Barred to practice their
profession for 7 years
• Third physicist is acquitted
26
28 patients treated with incorrect doses
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”
•
•
27
27
7/20/2005
25 July 2005
7
Components of 3-D TPS
Components of 3-D TPS
Hardware
•
•
•
•
•
•
•
•
•
•
28
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)]
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
28
7/20/2005
25 July 2005
29
Components of 3-D TPS
Algorithm Classification
Software 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
30
30
7/20/2005
25 July 2005
29
7/20/2005
25 July 2005
•
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
•
•
31
Imaging geometry and density data
Assumptions of symmetry
31
7/20/2005
25 July 2005
8
A. Scatter Integral
B. Use of Anatomy Data
Superposition Principle
Beam
Kernel
Slab
Kernel
Pencil
Kernel
Point
Kernel
•
Patient’s Anatomy
•
As imaged by CT, MR, PET, etc
Geometry and density
•
Symmetry assumptions
•
•
•
32
32
7/20/2005
25 July 2005
33
If you are buying...
Does the algorithm “feel” the
anatomy voxels in 1-D? 2-D? 3-D?
• Is the algorithm
•
As sensed by algorithm
1-D, 2-D, 2.5-D, or 3-D matrix
33
7/20/2005
25 July 2005
http://www.opctonline.net/
- Interesting background information
Of course
we use
a 3-D
convolution!
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?
•
34
34
7/20/2005
25 July 2005
35
35
7/20/2005
25 July 2005
9
http://www.opctonline.net/
36
36
7/20/2005
25 July 2005
37
Components of QA Program
37
7/20/2005
25 July 2005
Commissioning and QA of TPS
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
•
User
Manufacturer
User
•
38
38
7/20/2005
25 July 2005
Input
Data
(Radiation,
Patient)
Output
Radiation
Therapy
Planning
System
Commissioning: Initial tests
Quality control: Reproducibility
39
Dose
Distributions
Compare to
expected
results
39
7/20/2005
25 July 2005
10
Regions of Different Dose
Calculation Accuracy
Sample Criteria of Acceptability
IAEA TRS 430
AAPM TG53
40
40
7/20/2005
25 July 2005
41
Acceptance Testing
•
•
•
•
•
Hardware
Software
Elekta
6, 10, 18 MV
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
•
42
•
Catalogue delivered components
•
•
Type Tests
What happens in reality!
•
41
7/20/2005
25 July 2005
Results compared to results of type tests
42
7/20/2005
25 July 2005
Venselaar & Welleweerd
Radioth Oncol 60: 203-213,
2001.
43
43
7/20/2005
25 July 2005
11
Commissioning
•
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
•
•
IAEA TRS-430 provides sample tests
•
•
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
•
System set-up/machine configuration
Patient anatomical representation
External beam commissioning
Brachytherapy commissioning
Plan evaluation tools
Plan output and data transfer
Overall clinical tests
•
44
44
7/20/2005
25 July 2005
45
45
7/20/2005
25 July 2005
Phantoms
•
IMRT & 3-D QA
Med-Tec
IAEA TRS 430
46
46
7/20/2005
25 July 2005
47
47
7/20/2005
25 July 2005
12
QA of the Non-Dosimetric
Components of a TPS: QUASAR®
Non-Dosimetric Issues
Image acquisition and transfer
Beam display
• CT image reconstructions
•
Disclosure
•
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.
48
3-D display
Automatic tools - autocontouring, automargin, etc
• Dose volume histograms
•
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.
48
7/20/2005
25 July 2005
•
49
Phantom Schematic
50
50
7/20/2005
25 July 2005
Anatomical volumes
•
Oncol. Biol. Phys., 44, 955-966, 1999.
•
Multiplanar CT image reconstructions
Digitally reconstructed radiographs
CT numbers to electron density conversion
49
7/20/2005
25 July 2005
Commercial Version of QA Phantom
51
7/20/2005
25 July Modus
2005
Modus
Medical Devices, London, Ontario
51
13
Reconstructed Image Verification
Oblique CT reconstruction
52
52
7/20/2005
25 July 2005
53
Digitally reconstructed radiograph
53
7/20/2005
25 July 2005
Multi-Institution Evaluation
•
2cm
Phantom used to evaluate 3 TPSs and 1 CT
simulator
Picker ACQSIM
Varian CADPlan
• ADAC Pinnacle
• Theratronics TheraplanPlus
•
•
54
54
7/20/2005
25 July 2005
5cm
55
55
7/20/2005
25 July 2005
14
)
%
o
l
y
s
t
1
ro
el
ny
D
es
C
y
t
Mo
ry
el
es
in
ne
ae
d
eC
s
ru
b
uV
ro
e
1
0
8
0
6
0
4
0
2
0
0
l
0
yP
150
140
130
120
110
e
(
3m
)
u
e
P
2
l
m
o
u
V
B
C
A
D
B
C
D
C
M
o
A
2
u
0
c
i
t
e
6
4
2
0
A
iC
0
0
0
r u
b
We
e
d
g
e
A
B
C
0
A
a
x
u
u
l
0
1
2
0
a
e
r
t
e
d
d
0
4
0
6
B
i
C
-4
m
8
0
o
10
s
e
0
(
%
)
D
4u
%
m
v
5
5
7
72
/
5
a
2
0
J
u
/
r
2
l
0
y
0
2
5
0
0
i
a
5
t
i
o
n
e
r
M
2
D
D
c
s
0
0
0
l
a
V
L
4
a
e
T
o
0
0
5
0
0
1
0
0
0
5
0
S
S
S
B
0
1
T
N
2
C
N
v
n
u
m
b
C
C
y
y
y
a
87
5
/
2
J
m
m
m
s
eL
s
r C
i
t
o
A
B
C
t
a
a
n
d
B
r
o
0
1
0
n
s
k
i
0
0
0
.
0
00
.
15
0
u
/
l
2
y
0
0
2
.
e
10
l
.
a
25
t
.
20
i
.
v
5
e
E
l
e
c
5
0
t
5
0
5
5
92
97
5
r
/
o
2
J
0
u
n
/
l
2
y
V o
r E
e
n
l
D
0
0
2
e
• S
• V
•n B
A
c
r
A
i
r
y
l
l
• V
L
• E
• R
R
5
e
e
e
i
bM
0
5
-
82
t
t
t
t
m
r
0
-
5
s
s
s
t
u
e
a
e
l
a
s
r
a
e
m d -Mi
i
e
a
r
ri m t a-M
i
y
i
f
k
i
a
t
cL o
m
i
i cL
n
a
rC n
e
n
a
n
n rC l o
5
0
0
o i
5
15
a
MLC Phantom
Multi-Observer Test
AAPM TG 66, 2003
Does leaf end align with phantom geometry
(air/acrylic interface)?
• Errors ≥ 2 mm,
identified 100% of
the time
• 1 mm errors were
identified 80% of the
time
60
60
7/20/2005
25 July 2005
61
Non-Dosimetric Components
•
•
61
7/20/2005
25 July 2005
Quality Control
PS = Patient specific, W = Weekly, M =Monthly,
Q = Quarterly, A = Annually,
U = After software or hardware update
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
•
62
62
7/20/2005
25 July 2005
63
63
7/20/2005
25 July 2005
1
Sonic digitizer, 2 Electromagnetic digitizer
IAEA TRS-430 Table 61
16
QA Administration
Summary
•
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
•
64
64
7/20/2005
25 July 2005
•
•
Process QA
Incident/error rate
Number of replans
• Timeliness
• Physician satisfaction
•
65
65
7/20/2005
25 July 2005
Key Issues
•
Acceptance and commissioning will take time and
effort
Usually considers dose calculation issues
• Should also includes imaging issues
66
Documentation
Verification
Communication
Hardware
Software
•
•
Education
TPS consists many components
•
•
User training
Well-defined (re)commissioning tests
Well-defined repeatability checks
Appropriate actions as needed
Documentation of results
Patient specific QC
•
Take Home Message
•
Formal QC program includes:
QA of non-dose components is also important
66
7/20/2005
25 July 2005
67
67
7/20/2005
25 July 2005
17
Summary
•
Quality assurance of radiation treatment planning
systems is a significant challenge
•
68
but not insurmountable!
68
7/20/2005
25 July 2005
18