Serial Tomotherapy
D.A. Low
Slide 1
___________________________________
___________________________________
Serial Tomotherapy
Daniel A. Low, Ph.D.
Mallinckrodt Institute of Radiology
Washington University School of
Medicine
St. Louis, Missouri
Slide 2
Outline
• Description of Tomotherapy
• Clinical Implementation
– Commissioning
– Clinical Setup and Treatment
– Patient-Specific QA
• Clinical Issues
– Abutment-region Dosimetry
– Superficial Doses
– Whole-Body Doses
– Room Shielding
Slide 3
___________________________________
___________________________________
___________________________________
___________________________________
___________________________________
___________________________________
___________________________________
___________________________________
___________________________________
___________________________________
___________________________________
___________________________________
___________________________________
Commercial Application
• NOMOS Corporation
–
–
–
–
Multileaf collimator (MIMiC)
“Indexing” hardware
Treatment planning software
QA tools
• >40 users worldwide
___________________________________
___________________________________
___________________________________
___________________________________
___________________________________
___________________________________
Serial Tomotherapy
D.A. Low
Slide 4
MIMiC
• 20 pairs of leaves = 20 cm diameter cylinder
• Binary operation (pneumatic)
• Leaves subtend 0.84 (“1 cm”) or 1.7 cm “(2 cm”)
width
– Longer targets treated in successive abutted slices
(indexes)
•
•
•
•
•
Fluence modulated during arc
Arcs subdivided into 5 degree bins
Fluence nearly arbitrary each bin
Up to 72 independent coplanar beams
Patient indexed between slices
Slide 5
___________________________________
___________________________________
___________________________________
___________________________________
___________________________________
___________________________________
___________________________________
___________________________________
Crane
• Couch repositioning and immobilization
• Attaches to couch rail
• Position read out using linear digital
encoders
___________________________________
___________________________________
___________________________________
___________________________________
___________________________________
___________________________________
Slide 6
___________________________________
Treatment Planning System
•
•
•
•
•
•
•
Trade-name Corvus
Structure delineation/import
DVH-based optimization user interface
Simulated annealing algorithm
Multiple 2-D cross-section views
DVHs
Also supports DMLC
___________________________________
___________________________________
___________________________________
___________________________________
___________________________________
___________________________________
Serial Tomotherapy
D.A. Low
Slide 7
___________________________________
Clinical Implementation
• Commissioning
– Traditional
– Dosimetric
• Patient setup and treatment
–
–
–
–
Immobilization
CT scan acquisition
Contour delineation and dose optimization
Treatment
• Patient-based QA
– Dosimetric pre-treatment
– Positioning verification
Slide 8
___________________________________
___________________________________
___________________________________
___________________________________
___________________________________
___________________________________
___________________________________
Commissioning
• Traditional, e.g.
–
–
–
–
TG 53
Patient information
Printout accuracy
Data transfer
• IMRT
– Dose distributions
– Monitor unit determinations
Slide 9
___________________________________
___________________________________
___________________________________
___________________________________
___________________________________
___________________________________
___________________________________
IMRT Dosimetry
• Dosimeters
– Accuracy and dimensionality conflict
– Point
•
•
•
•
Ionization chambers
TLD chips
Others
Poor spatial coverage
– Size vs. dose heterogeneity
– Response as function of incident fluence angle
___________________________________
___________________________________
___________________________________
___________________________________
___________________________________
___________________________________
Serial Tomotherapy
D.A. Low
Slide 10
___________________________________
IMRT Dosimetry
• Dosimeters (cont’d)
– Planar
• Radiographic Film
– Poor energy response
– Convenient, inexpensive, large format
• Radiochromic Film
– Still difficult to use quantitatively
– Possible to get 2% precision with 0.1 x 0.1 mm2
resolution!!
– Excellent for benchmarking
Slide 11
___________________________________
___________________________________
___________________________________
___________________________________
___________________________________
___________________________________
___________________________________
IMRT Dosimetry
• Dosimeters (cont’d)
– Volumetric
•
•
•
•
Polymerizing gel (Trade name BANG)
MRI or optical readout
Reasonable sensitivity (1-20 Gy)
MRI readout shown to be excellent as relative
dosimeter (single experiment yields nearly
1,000,000 1 x 1 x 3 mm3 data points!)
• MRI readout still requires benchmarking for
absolute IMRT dose measurements
• Optical readout requires benchmarking for large
volumes
Slide 12
___________________________________
___________________________________
___________________________________
___________________________________
___________________________________
___________________________________
___________________________________
IMRT Dosimetry
• Phantoms
– Anthropomorphic
• Geometrically irregular
• Patient-like structures including heterogeneities
• Heterogeneities may yield unnecessary dosimetric
uncertainties
• Highly accurate spatial film registration difficult
___________________________________
___________________________________
___________________________________
___________________________________
___________________________________
___________________________________
Serial Tomotherapy
D.A. Low
Slide 13
___________________________________
IMRT Dosimetry
• Phantoms (cont’d)
– Geometrically regular
•
•
•
•
Easily aligned and registered
Precise internal construction
Homogeneous internal construction
Multiple dosimeters measuring in same dose
distribution environment
___________________________________
___________________________________
___________________________________
___________________________________
___________________________________
___________________________________
Slide 14
___________________________________
Patient Setup and Treatment
• Immobilization
– Dose delivered in sequential slices
– If patient moves perpendicular to arc during or
between slice delivery, large dose heterogeneity
results in abutment region
– Immobilization system should concentrate on
longitudinal direction
– Don’t forget localization to enable smallest
possible margins
Slide 15
Patient Setup and Treatment
• Volumetric anatomy measurement
– Usually done with CT
– Targets and critical structures delineated
• Optimization
– Commercial system uses DVH-based system
– Target and critical structure dose limits entered as 3point DVHs
– Pre-filtering of gantry angles
– Simulated annealing determines individual beam
fluences
– Patient treatment plan written on a floppy disk which
monitors number of delivered fractions
___________________________________
___________________________________
___________________________________
___________________________________
___________________________________
___________________________________
___________________________________
___________________________________
___________________________________
___________________________________
___________________________________
___________________________________
___________________________________
Serial Tomotherapy
D.A. Low
Slide 16
___________________________________
Patient Setup and Treatment
• Laser alignment marks placed similar to
conventional CT
• Intersection of alignment marks determines
coordinate system “Origin”
• Gantry rotation axis usually passes through
isocenter, but tool is provided to move axis
if necessary
• Patient is aligned to marks, CRANE
readouts zeroed
• Patient moved as required by treatment plan
Slide 17
Patient Setup and Treatment
___________________________________
___________________________________
___________________________________
___________________________________
___________________________________
___________________________________
___________________________________
___________________________________
• Accelerator uses normal arc mode
• MIMiC leaves open and close as function of
gantry angle (inclinometers)
• MIMiC controlling computer monitors
gantry rotation rate and angles to determine
of treatment proceeding correctly
• Monitor units or dose are not monitored by
MIMiC controlling computer
___________________________________
___________________________________
___________________________________
___________________________________
___________________________________
Slide 18
___________________________________
Patient-Based QA
•
•
•
•
•
Treatment plan yields monitor units
For same dose, MUs can vary by 40%
How do we know the MUs are correct?
MEASUREMENT!
Patient fluence distribution transferred to
QA phantom and dose measured at select
points
• Ionization chambers for MU normalization
• Film used for spatial localization
___________________________________
___________________________________
___________________________________
___________________________________
___________________________________
___________________________________
Serial Tomotherapy
D.A. Low
Slide 19
___________________________________
Patient-Based QA
___________________________________
• Patient Localization QA
– How do we know the patient is in the correct
location?
– PORTAL FILMS
– Compared against DRRs from commercial
virtual simulation software using same CT
dataset and radiopaque markers as treatment
plan
– Use double exposures when possible (with and
without MIMiC)
Slide 20
___________________________________
___________________________________
___________________________________
___________________________________
___________________________________
___________________________________
Clinical Issues
• Abutment-Region Dosimetry
– CRANE indexing precision
– Dosimetric consequences of indexing error
– Intrinsic abutment dose distribution
heterogeneity
___________________________________
___________________________________
___________________________________
___________________________________
___________________________________
___________________________________
Slide 21
___________________________________
CRANE Indexing Precision
• Readout precision = 0.01 mm
• Readout made at rack and pinion gears, not at
accelerator isocenter
• Couch bearing friction limits precision
• Measurement:
– Radiographic using sequence of open MIMiC field images
– Purposefully change index amount from overlap to underlap
– Scan film and correlate hot and cold spots to intended couch index
movement
– Fabricate high precision system to provide baseline for relationship
between hot and cold spots and couch index distance (“gold
standard”
– Also evaluate optical-based system: miniCRANE
___________________________________
___________________________________
___________________________________
___________________________________
___________________________________
___________________________________
Serial Tomotherapy
D.A. Low
Slide 22
16.3 mm
16.8 mm
17.2 mm
___________________________________
___________________________________
___________________________________
___________________________________
___________________________________
___________________________________
___________________________________
Slide 23
Gold Standard
CRANE
30
Dose Error (%)
20
___________________________________
___________________________________
Square = gold standard
Triangle = miniCRANE
Circle = CRANE
10
___________________________________
___________________________________
0
___________________________________
−10
___________________________________
−0.5
CRANE
0
0.5
Index offset (mm)Mini CRANE
Slide 24
___________________________________
___________________________________
Indexing Precision Results
• Standard deviation
– CRANE = 0.10mm
– miniCRANE = 0.08 mm
– gold standard = 0.02 mm
___________________________________
___________________________________
___________________________________
___________________________________
___________________________________
___________________________________
Serial Tomotherapy
D.A. Low
Slide 25
___________________________________
Consequences of Incorrect
Indexing
• Either incorrect indexing or patient
movement yields undesired overlap or
underlap between successive abutments
• Carol determined dose error 10% mm-1
• Measurement
– 8 cm diameter target
– Radiographic film (coronal)
– Purposeful index error 0, ±1, ±2 mm
Slide 26
___________________________________
___________________________________
___________________________________
___________________________________
___________________________________
___________________________________
___________________________________
-2 mm -1 mm 0 mm +1 mm+2 mm
___________________________________
___________________________________
___________________________________
___________________________________
___________________________________
___________________________________
Slide 27
___________________________________
Dose Error vs Index Error
60
___________________________________
40
Dose Error (%)
Slope = 25%
mm-1
___________________________________
20
___________________________________
0
___________________________________
−20
−40
−60
−2
___________________________________
−1
0
1
Offset from 16.8 cm (mm)
2
___________________________________
Serial Tomotherapy
D.A. Low
Slide 28
___________________________________
Intrinsic Abutment Dosimetry
___________________________________
• Narrow but divergent beams
–
–
–
–
Unequal matching at abutments
Hot and cold spots created
Only narrow width
Dose heterogeneities are function of gantry
angle rotation, off-axis distance, width of leaves
(“1 cm” vs “2 cm” modes)
– Random daily setup error may redistribute and
reduce dose heterogeneity magnitude
Slide 29
Intrinsic Dose Heterogeneity
Measurement
• Treatment Plans
–
–
–
–
___________________________________
8 cm cylindrical targets (head phantom)
Position relative to isocenter
180°, 240°, 300° arcs
“1” and “2” cm modes
• Measurement
– Radiographic film (coronal)
– Precise indexing of phantom
– Densitometry - 0.25 mm laser digitizer
Slide 30
___________________________________
___________________________________
___________________________________
___________________________________
___________________________________
___________________________________
___________________________________
___________________________________
___________________________________
___________________________________
___________________________________
___________________________________
Experimental Layout
___________________________________
Isocenter
MIMiC Projection
10 cm radius
___________________________________
Target Volume
4 cm radius
Extrapolated
Region
___________________________________
180º
180º
x
___________________________________
240º
240º
___________________________________
300º
300º
___________________________________
y
Serial Tomotherapy
D.A. Low
Slide 31
___________________________________
Experimental Method
___________________________________
• Determine heterogeneity throughout 20 cm
diameter volume
___________________________________
– Apply smooth fit in x and y
• Position of abutment region in patient
–
–
–
–
Slide 32
___________________________________
Longitudinal position has random error
Smooths (distributes) abutment region
Model as Gaussian distribution
Convolve with abutment hot/cold spots
___________________________________
___________________________________
Measured
Overlap Profiles
1
Dose (arb. units)
0.8
0.6
↑
300 °
↑
240 °
___________________________________
___________________________________
___________________________________
___________________________________
↑
180 °
Abutment Example
___________________________________
0.4
___________________________________
0.2
___________________________________
0
0
50
100
z (mm)
Slide 33
150
___________________________________
___________________________________
180°, 1 cm
___________________________________
___________________________________
___________________________________
___________________________________
___________________________________
___________________________________
Serial Tomotherapy
D.A. Low
Slide 34
___________________________________
180°, 1 cm Example
___________________________________
___________________________________
___________________________________
___________________________________
___________________________________
___________________________________
Slide 35
___________________________________
Result Presentation
___________________________________
___________________________________
• 2D contour plot difficult to interpret
• Largest variation in “y” direction
• Show 1-D plots of smoothed dose
heterogeneity in “y” direction with
superimposed measured data points
___________________________________
___________________________________
___________________________________
___________________________________
Slide 36
___________________________________
180° arc, 1cm mode
___________________________________
30
Dose Heterogeneity (%)
20
___________________________________
10
___________________________________
0
−10
___________________________________
−20
−30
−40
−50
100
σ = 0 mm
σ = 1 mm
σ = 2 mm
σ = 3 mm
50
___________________________________
0
y (mm)
−50
−100
___________________________________
Serial Tomotherapy
D.A. Low
Slide 37
___________________________________
180° arc, 2cm mode
___________________________________
60
Dose Heterogeneity (%)
40
___________________________________
20
___________________________________
0
−20
−40
−60
−80
100
___________________________________
σ = 0 mm
σ = 1 mm
σ = 2 mm
σ = 3 mm
50
___________________________________
0
y (mm)
−50
−100
Slide 38
___________________________________
___________________________________
240° arc, 1cm mode
___________________________________
Dose Heterogeneity (%)
20
10
___________________________________
0
___________________________________
−10
___________________________________
−20
−30
−40
100
σ = 0 mm
σ = 1 mm
σ = 2 mm
σ = 3 mm
50
___________________________________
0
y (mm)
−50
−100
Slide 39
___________________________________
___________________________________
300° arc, 1cm mode
___________________________________
15
Dose Heterogeneity (%)
10
___________________________________
5
0
___________________________________
−5
___________________________________
−10
−15
−20
−25
100
σ = 0 mm
σ = 1 mm
σ = 2 mm
σ = 3 mm
50
___________________________________
0
y (mm)
−50
−100
___________________________________
Serial Tomotherapy
D.A. Low
Slide 40
___________________________________
300° arc, 2cm mode
___________________________________
Dose Heterogeneity (%)
20
0
___________________________________
−10
___________________________________
−20
−30
−40
100
Slide 41
___________________________________
10
σ = 0 mm
σ = 1 mm
σ = 2 mm
σ = 3 mm
50
___________________________________
0
y (mm)
−50
−100
Intrinsic Abutment Summary
Target off-axis distance along y axis required to achieve 10% dose
homogeneity
1 cm mode
σ=0mm
σ=2mm
ant post
ant post
300º 5
10
8
10
180º 3
3
5
6
2 cm mode
ant post
ant post
300º 3
6
5
10
180º 1
2
3
4
Slide 42
___________________________________
___________________________________
___________________________________
___________________________________
___________________________________
___________________________________
___________________________________
___________________________________
___________________________________
Recommendations
• 2 cm mode yields approx 70% greater
heterogeneities than 1 cm, use 1 cm
whenever practical
• Keep targets near isocenter
• Use as large a gantry angle as possible
• Periodically monitor couch indexing
precision
___________________________________
___________________________________
___________________________________
___________________________________
___________________________________
___________________________________
Serial Tomotherapy
D.A. Low
Slide 43
___________________________________
Future Implementation
___________________________________
• Spiral tomotherapy unit - Mackie
• Abutment region heterogeneities distributed
throughout patient
• Improved patient throughput
• On-line images acquired during irradiation
may yield tomographic information
___________________________________
___________________________________
___________________________________
___________________________________
___________________________________
Slide 44
___________________________________
Acknowledgement
• This work was supported in part by a grant
from the NOMOS corporation
___________________________________
___________________________________
___________________________________
___________________________________
___________________________________
___________________________________