Uniform Scanning and Energy Stacking with Proton Beams AAPM - - PowerPoint PPT Presentation

Uniform Scanning and Energy Stacking with Proton Beams

AAPM Continuing Education Session 22-Jul-2010

Outline

 Introduction to Technique - Moyers (15 min)

» description of delivery techniques and terminology » radiobiology » lessons from scanning electron beam incidents » advantages and disadvantages of technique

 Design and Implementation of Safe Delivery Systems - Anferov (15 min)

» potential hazards » example hazard mitigations

 Practical Aspects - Hsi (15 min)

» optimization of scan and stacking patterns » multi-element detectors for measurements » scanning and stacking specific QA

 Questions - All (10 min)

Scanning Terminology

 scanning modes (as defined by DICOM-RT ion)

» none » uniform scanning » modulated scanning

 repainting  uniform scanning patterns

» Lissajous » circular (single or multiple) » raster (rectilinear) » spiral » triangle

History of Uniform Scanning in the Clinic

 Early

» Michael Reese e patient scan  1955 » Uppsala p Lissajous beam scan 1957 » Sagittaire e Lissajous beam scan  1970 » Berkeley He, Ne raster and circular beam scans  1985

 Recent

» Mitsubishi p, C circular beam scan  1995 » IUCF/MPRI p raster beam scan  2005 » IBA p triangle beam scan  2007 » Mitsubishi p spiral beam scan  2011 » Sumitomo p spiral beam scan  2011

Uniform Dose Coverage of Target in Depth Direction

 energy

stacking

 rotating

propellors

 ridge

filters

 stepped

cones

Energy Stacking Methods

 direct extraction from accelerator  rangeshifter near accelerator  rangeshifter near gantry  rangeshifter in radiation head

Rangeshifter Types

 binary slabs linear double wedges circular double wedges  circular steps

Number of Requestable Energies

Example Method Comments synchrotron interpolation 18,000 energies between 70 and 250 MeV synchrotron pre-programmed 256 energies (approximately 1 mm steps) cyclotron RS in SY 1 mm range steps cyclotron RS in head  30 range steps

Energy/Range Stability - During Treatment

 accelerator

energy stability

 RS

thickness stability

 patient

thickness stability

Energy/Range Reproducibility - Day-to-Day

 accelerator

energy reprod.

 RS

thickness reprod.

note green area represents  0.18 mm

Energy Switching Time

 variable energy synchrotron

» could change energy multiple times during each cycle » typically only a single energy is extracted each cycle » verify energy before extraction » need to reconfigure SY

 fixed energy cyclotron

» move RS ( 0.2 s) » verify energy before delivery » for SY RS, need to reconfigure SY » for head RS, do not need to reconfigure SY

MU Considerations

 Uniform dose coverage concerns

» flux non-uniformity during delivery » shifting scanning patterns » starting or stopping beam delivery in middle of scan pattern

 Better dose uniformity with integral

number of repaintings and larger number of repaintings.

 Shallow layers use a small fraction of

the total MU; difficult to repaint.

 Flux rate, scan pattern, scan speed,

and number of repaintings must be carefully balanced.

 Typically the MU per portal is restricted

to a minimum value.

Das et al., (1994)

Interplay with Patient Motion

 motion of beam versus

motion of patient

» if a person walks back and forth through a scattering sprinkler, will get a little wet » if a person walks back and forth through a scanning sprinkler, may stay dry or may get soaked  fast uniform scanning of

each layer is typically faster than respiration but slow energy stacking may be an issue. scanning sprinkler  scattering sprinkler 

Radiobiology of Scanned Beams

 thus far no direct comparison

• f uniform scanned proton

beams to scattered proton beams

 experiments with scanned

electron beams showed RBE up to 1.29 depending upon scan pattern

Meyn et al., (1991)

Lessons Learned from Previous Incidents with Scanned Electron Beams

 Sagittaire example

» bending magnet power supply stuck at wrong high energy (32 MeV) » energy feedback loop adjusted energy so beam would pass through energy analyzing slit in bending magnet stuck at wrong high energy » scanning magnet power supply set at correct low energy (13 MeV) » upstream dose monitor measured correct whole beam flux but fluence distribution downstream concentrated in middle of field » one patient had parallel opposed posterior cervical strip fields resulting in  800 cGy to spinal cord in one fraction » medical problems for patient within 45 minutes

 Lessons for safety

» energy interlocks for accidents - DAILY QA OF THE RANGE IS NOT SUFFICIENT » downstream fluence distribution detectors

Advantages and Disadvantages

 advantages

» uniform dose distribution for all energies and field sizes » smaller loss of range for large fields compared to scatterer technique » no need for electromechanical scatterer exchangers » higher particle use efficiency / less neutron production

 disadvantages

» requires additional time to switch energies » minimum MU constraint for portal » increased interplay with patient motion » requires diligent safety system

Part 2

Making Uniform Scanning safe

Safe Design Practices

 If it can break – it will

i.e consider all failure modes and look at the outcome  Failure Modes & Effects Analysis (FMEA) process:

» Define failure modes and associated risks » Add mitigations that

– Reduce probability of a failure mode – Detect failure and stop before any harm is done

 Use KISS principle:

» Keep It Simple Stupid !

Uniform Scanning Features



Insensitive to beam misalignment



High instantaneous dose rate



Beam spot size can vary from 0.5 to 1.0 Line Spacing without perturbing uniformity



Scan pattern can be started and validated prior to delivering dose to the patient !

1 10 100 1000 Dose Rate (Gy/min) Double Scattering Uniform Scanning Spot Scanning

Dose rate in a beam spot for average 2Gy/min to 10

3 cm 3

1 2 3 4 5 6 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2 2.2 Line Spacing / Sigma Overscan/ 1 2 3 4 5 6 Ripple [%] Overscan Ripple

Hazard Ratings

 No perceptible effect  Small loss of performance  Loss of product function, but no damage to

user, patient, equipment.

 Possible injury without irreversible damage  Possible injury with permanent damage  Death of user or patient MINOR MODERATE HIGH CRITICAL Insignificant Catastrophic

New Hazards due to Scanning System

• A. High dose rate in a beam spot can cause

large dose errors if scanning stops

» 5% dose error can accumulate in 5 msec. » 100% dose error can accumulate in 100msec

• B. Non-uniform transverse dose distribution due

to errors in the scanning pattern.

» Accumulates over the course of the treatment.

Critical High

Sensitivity to Beam Failures

strong weak*

• 7. Beam Energy / per layer

strong weak

• 6. Intensity Fluctuation

moderate weak

• 5. Beam Intensity Rate

moderate n/a

• 4. Beam spot halo

strong weak

• 3. Beam spot shape

strong weak

• 2. Beam spot size error

strong weak

• 1. Beam misalignment

Spot Scanning Uniform Scanning

Beam Failures

* Only if using passive range modulation (ridge filters, range shifter in the nozzle)

Safety Mitigations

 Start scanning verification prior to dose delivery

» Apply checks that validate scan profile, scan amplitudes and scan accuracy.

 Perform scanning system health validation at a fast rate

(~1kHz) and interlock beam delivery.

» Redundant hardware checking mitigates critical hazard of burning a hole through the target.

 Monitor Field Flatness, Size and Symmetry throughout the

treatment using segmented ion chamber.

» This check validates accuracy of the dose delivery process.

Scanning System Health Checks

A

Generator Every 0.1 ms must receive a trigger pulse indicating Generator updated its output

A,B

Magnet Magnet health: analog circuit monitors voltage from the pickup coils.

B

Generator, Magnet Waveform stability: waveform parameters do not change during treatment

B

Generator, waveform Measure waveform parameters: Min and Max values of Currents, Voltages, Frequencies

B

Power Supply errors Every 1 ms PS output is within tolerance from Generator

A

Generator or Power Supply Every 1 ms PS output change indicate beam spot motion 1 cm or more Hazard Failure Mode Hardware Health checks

Scanning is only part of the picture

 Treatment energy setup

validate beam penetration range

 Lateral Beam spreading

validate scanning safety

 Dose modulation in depth

validate ridge filters / range shifters

 Dose conformation to target validate collimator & bolus  Measure the dose

Redundant dose counters, MUs agreement

Dose delivery system

Safety Checks

Safety Summary

 Compared to a Double scattering system

Uniform scanning adds two new hazards:

» Stopped scanning » Incorrectly executed scanning

 With dedicated safety electronics monitoring health of the

scanning system uniform scanning can be safe and robust alternative to both double scattering and pencil beam scanning

Part 3

Practical aspects

– utilize uniform

scanning & discrete energy stacking protons for treatments

Maglev train at China with maximum speed of 431 km/h (268 mph)

Virtual source distances for various scanning magnets

Combined scanning magnet

» single virtual source distance

Dual scanning magnets

» two virtual source distances » effective source size and distance are same for both axes

Parallel scanning

» single "infinite“ virtual source distance

Requirements of clinical performance

Dose Reference Volume (DRV)

• Transverse – lateral extent

Inside 2-times penumbra Penumbra width

• Depth –longitudinal extent

 Within modulation width Requirements for lateral extent at above are only applied for depths with the longitudinal extent.

Scan Pattern Optimization - spot shape & size in air

Spot shape in air, i.e. proton fluence distribution at the entrance of patient.

Large section near patient is under vacuum for IBA delicated nozzle at Essen. Measured spot sizes for this nozzle are ~4 and 6mm for beam ranges of 32 and 20 cm in circular

• shape. However, when a beam-position

monitor located at entrance of gantry was inserted as extra material into beamline, elongated beam spot was observed as shown at most left of top panel. Spot sizes in air also measured for a universal nozzle at MPRI are 6 to 14mm for various ranges as shown at bottom panel. Larger spot size for universal nozzle is due to no applied vacuum from scanning magnet to patient.

Scan Pattern Optimization - spot size in patient

Spot sizes in air σair measured at MPRI was fit as a function of range R; shown dash line. Spot sizes σpt due to scatters in patient is calculated by 0.02275R + 0.12085E-4R2 as in Hong et al publication as open circles. Total spot size including initial spot size and scatters in patient as close circles is calculated by for each beam range σtot (R) = (σair

2+ σpt 2)1/2

Scan Pattern Optimization - Path spacing & over-scan inside field

Optimization based path spacing & over-scan inside field

Effects of collimation - Over-scan beyond field edge

Over-scan distance beyond edges of beam-limiting devices

• upstream trimming collimators & patient aperture

Scan Pattern Optimization - various depths

Scan pattern optimization of spot size and path spacing focus

• n depths within longitudinal extent and the center of

modulated protons at beamline isocenter.

Energy Stacking Optimization - Depth Spacing

Widths of non-modulated Bragg-Peak (BP) depth doses

varies with proton energy (i.e. beam range) and delivery system

3 options for stacking energy layers to form required flatness over longitudinal extent are needed for MPRI TR2. For options with ranges

• f 12-20cm and 20-27m, depth

spacing between consequent layers is 0.6 cm. However, 0.3 cm depth spacing is required for ranges of 4-12 cm when the width is only 0.5 mm for 4 cm range. For widths from 3cm to 1.5cm, 4 options are needed for OKC-IBA US beam-line with ~0.6cm depth spacing for all beam ranges.

Energy Stacking Optimization - Weights of energy layers

After depth-spacing is chosen for each option,

• riginal weights of energy layers were obtained

for standard range of 16cm by theoretical model as shown blue points. During commission, depth doses with original weight were measured as shown blue points at bottom panel. A correcting algorithm was used to adjust ~8% title. Optimized weights of energy layers were then

• btained as shown red points in both panels.

Because widths of non-modulated protons varies significant between options, using optimized weights from standard options results significant tilt on depth doses for non-standard options. The correcting algorithm described above is used for obtaining optimized weights for various

• ptions during commission. Similar procedure

has been internally performed by IBA vendor.

Energy Stacking Optimization - between proton delivery systems

Weights of energy layers

• Energy options – range-shifter thickness and MU courting per proton
• Scattering generated by materials used in energy stacking

Weights

• f
• ptimized

energy- stacking layers vary between energy options; depend on range- shifter thickness and MU courting per proton. Although the trend of weights used energy layers is similar between different proton beam-lines, subtle difference can be related scattering generated by energy layers in different beam- lines.

Depth Dose Measurements

• Use of a single chamber to measure the dose at all depths requires repeating

the whole delivery sequence for each depth.

• Use of a MLIC (multiple layer ionization chamber) allows measurement of

dose at many depths using a single delivery.

10 20 30 40 50 60 70 80 90 100

3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28

Depth [cm] DD [%]

R 27cm Scan 3cm Circular Pt-by-Pt 3cmcircular MLIC F.S. 10cm Pt-by-Pt, F.S. 10cm MLIC F.S. 10.cm

Dmitri et. al. 2007

Lateral Dose Profile Measurements

Film Step-by-step method with mini-chamber MPIC (multiple pad ionization chamber) - 1D

• r 2D configuration

In-plane Profiles

10 20 30 40 50 60 70 80 90 100 110

• 8
• 7
• 6
• 5
• 4
• 3
• 2
• 1

1 2 3 4 5 6 7 8

X (cm) Profile (%)

Water Phantom MPIC

Treatment failure recover

Partial treatment delivery

Routine QA - Range-shifter of energy layer

Check beam range at off-axis positions to verify constancy of range-shifter material thickness used for energy stacking.

Routine Modulation QA

Ensure that files storing weights of energy layers for various modulations of each option are not corrupted.

» Check weights of energy layers for standard condition daily. » measure depth dose distributions monthly. » compare routinely used files with secure master files annually.

For a proton system that weight and scanning aptitude of each energy layer are determined by an algorithm as function of beam range and modulation, ensure that the calculation algorithm is not corrupted

» Check weights of energy layers for standard condition daily » measure depth dose distributions and lateral profiles monthly.

Summary of practical aspects for utilizing uniform-scanning protons

• Optimizations of scanning pattern and energy-stacking need be

performed to satisfy clinic requirements on lateral and longitudinal extent for providing good treatments.

• Specific dosimeters for depth doses and lateral profiles are needed

for commissioning a US proton beamline efficiently.

• Recovery treatment for partial delivery is required.
• Beam range verify at off-axis positions is necessary when large

area of range-shifter is used for energy stacking during commission.

• Routine QA for off-axis range constancy as hardware and

modulation as software should be performed to assure correct dose delivery.

References



Moyers, M. F. “Proton Therapy” The Modern Technology of Radiation Oncology: A Compendium for Medical Physicists and Radiation Oncologists ed. van Dyk, J. (Wisconsin: Medical Physics Publishing, 1999) p. 823 - 869.



Meyn, R. E. Peters, L. J. Mills, M. D. Moyers, M. F. Fields, R. S. Withers, H. R. Mason,

• K. A. "Radiobiological aspects of electron beams" Frontiers of Radiation Therapy and

Oncology 25 eds. Vaeth, J. M. and Meyer, J. L. (S. Karger AG Basel, Switzerland, 1991)

• p. 53 - 60.



Moyers, M. F. "LLUPTF: eleven years and beyond" Nuclear Physics in the 21st Century (New York: American Institute of Physics, 2002) p. 305 - 309.



Moyers, M. F. Vatnitsky, S. M. Practical Implementation of Light Ion Beam Treatments (Wisconsin: Medical Physics Publishing, 2011).



• V. Anferov, “Scan pattern optimization for uniform proton beam scanning”, Med. Phys.

36(8), 3560 (2009).



J.B. Farr et al., “Clinical characterization of a proton beam continuous uniform scanning system with dose layer stacking”, Med. Phys. 35(11), 4945 (2008).



Das, I. J. et al., "Dosimetric problems at low monitor unit settings for scanned and scattering foil electron beams" Med. Phys. 21(6), 821 (1994).



Nichiporov, D. et al., “Multichannel detectors for profile measurements in clinical proton fields”, Med. Phys. 34(7), 2683 (2007).

Questions

Interesting Stuff

http://www.ngsir.netfirms.com/englishhtm/Lissajous.htm