F. Sauli - ICPT Workshop (Oxford, 23-26 September 2008) 0 AQUA A DVANCED QU ALITY A SSURANCE FOR...

1

F. Sauli - ICPT Workshop (Oxford, 23-26 September 2008)

AQUA

ADVANCED QUALITY ASSURANCE FOR CNAO

AQUA

ADVANCED QUALITY ASSURANCE FOR CNAO

U. Amaldi, S. Iliescu, N. Malakhov, J. Samarati, F. Sauli and D. WattsTERA Foundation and CERN

Presented by Fabio Sauli

U. Amaldi, S. Iliescu, N. Malakhov, J. Samarati, F. Sauli and D. WattsTERA Foundation and CERN

Presented by Fabio Sauli

Advanced Instrumentation for Cancer Diagnosis and TreatmentESF workshop (Oxford 23-26 September 2008)

Advanced Instrumentation for Cancer Diagnosis and TreatmentESF workshop (Oxford 23-26 September 2008)

2


CNAO (CENTRO NAZIONALE DI ADROTERAPIA ONCOLOGICA)CNAO (CENTRO NAZIONALE DI ADROTERAPIA ONCOLOGICA)

132

NATIONAL CENTRE OF ONCOLOGICAL HADROTHERAPY (Pavia, Italy)

Synchrotron accelerator for light ions (protons, carbon) up to 400 MeV/u3 fixed beam treatment rooms, 2 gantries

Startup in spring 2009

Status of the accelerator (March “08):

3


AQUA PROGRAMS FOR CNAOAQUA PROGRAMS FOR CNAO

In-beam PETCrystalsResistive Plate Chambers

PRRProton Range Radiography

IVI Interaction Vertex Imaging

NSTNuclear Scattering Tomography

4


PRR: PROTON RANGE RADIOGRAPHYPRR: PROTON RANGE RADIOGRAPHY

K.M. Hanson et al, Phys. Med. Biol. 26 (1981) 965

P. Pemler et al, Nucl. Instr. and Meth. A432(1999)483

For beam energies above total absorption in the target, a correlated measurement of track position and residual energy allows to reconstruct the integrated density image:

The residual energy is commonly measured with a monolithic crystal scintillator. Alternative system use of a plastic scintillator stacks.

€

ΔE =dE

dl

⎛ ⎝ ⎜

⎞ ⎠ ⎟l0

L

∫ dl = ρ(l)S(l,E l )dl0

L

∫

ΔEE ER

L

€

σ R

R=

σ S

R

⎛ ⎝ ⎜

⎞ ⎠ ⎟2

+σ P

R

⎛ ⎝ ⎜

⎞ ⎠ ⎟2

+σ C

R

⎛ ⎝ ⎜

⎞ ⎠ ⎟2

Range resolution due to straggling (σS) and beam momentum spread (σP)

€

σ ρ

ρ=

σ R

L N

5


PROTON RANGE TELESCOPEPROTON RANGE TELESCOPE

TrackerGas Electron Multiplier (GEM) detectorsDigital Strip Readout

Range/Energy loss Plastic scintillator StackSilicon Photomultipliers Readout

Measurement of range and energy lossHigh intrinsic rate capability (>1 MHz)Density 1, (almost) tissue equivalentLow cost, easily scalable to larger sizes

Using modern technology developed for High Energy Physics

Light, 2-dimensional readoutHigh rate (> 1 MHz)Radiation resistantLarge areas at low cost

6


GEANT 4 SIMULATIONSGEANT 4 SIMULATIONS

Mean differential energy loss in 3 mm thick scintillator slices and projected range as a function of initial energy (50 to 200 MeV in 25 MeV steps:

50

100

150

200 MeV

Range straggling~ 1.5 % rms

7


RANGE DETERMINATION RANGE DETERMINATION

Due to straggling, two events with the same initial energy have a different energy loss profile, depending on penetration in the last scintillator slice:

A simple algorithm, the fraction of energy loss in the last scintillator to the one before the last, allows to deduce the fractional penetration in the last slice:

€

R = N − 1( ) + fAN

AN + AN−1

⎧ ⎨ ⎩

⎫ ⎬ ⎭T

AN : signal in last slice NAN-1 : signal in slice N-1T: slice thickness

8


ENERGY LOSS: ELECTROMAGNETIC VS NUCLEARENERGY LOSS: ELECTROMAGNETIC VS NUCLEAR

A fraction of events (~25%) undergo a nuclear interaction. The residual range for a given initial energy has a long tail at short values:

electromagnetic only

nuclear+electromagnetic

Electromagnetic

Nuclear

The expected range-energy correlation can be exploited to discard nuclear interaction events:

9


PRR RESOLUTION STUDYPRR RESOLUTION STUDY

In collaboration with Prof. George ChenMassachusetts General Hospital (Boston)

Simulated Proton Range Radiography, deduced from conventional CAT scans: Difference between the initial range map and

the subsequent map generated from the initial planning CT scan and the follow-up scan after 5 weeks of treatment:

GOALS:Treatment plan verificationOrgan motion correction

10


SCINTILLATOR RANGE TELESCOPESCINTILLATOR RANGE TELESCOPE30 scintillators 3 mm thick, 10x10 cm2, read-out with wavelength shifter fibres and solid state sensors

~ 16 photoelectrons

Fast scintillator: BC-408• Rise/decay time 0.9/2.1 ns• Peak emission 425 nmWLS: BC-482A 1mm • Decay time 12 ns• Absorption peak 420 nm• Emission peak 494 nmMMPC: Hamamatsu S-10362-11-050U• 1 mm2 active, 400 pixels• Quantum efficiency 40% at 500 nm• Time resolution 200-300 ps Response to minimum ionizing electrons

(protons are from 3 to 20 times MIPs)

11


SCINTILLATOR RANGE TELESCOPESCINTILLATOR RANGE TELESCOPEThe scintillator modules are mounted on a frame support; the space between counters permits the insertion of absorbers to extend the residual energy range.

Two scintillators back to back with readout electronics:

12


TRACKING DETECTORS FOR PRR TRACKING DETECTORS FOR PRR

F. Sauli, Nucl. Instr. and Meth. A386(1997)531

GEM: GAS ELECTRON MULTIPLIERSFast gaseous position-sensitive detector, used in HEP experiments. The ionization released in a thin gas layer by charged particles is amplified and detected on perpendicular readout strips.Typical performances:• Single particle detection and recording• Position accuracy ~ 100 µm• Rate capability > 1 MHz/cm2

• Radiation resistance above 1014 particles/cm2 • Low mass ~ 0.5% X0 per 2-D detector (0.3 mm H2O• TIME-STAMPED EVENTS

GEM chamber, 10x10 cm2 active Fast digital readout

Completion, test and calibration: fall ‘08Installation at CNAO: spring ‘09

13


GEM DETECTORS GEM DETECTORS

GAS ELECTRON MULTIPLIER

Typical geometry:5 µm Cu on 50 µm Kapton70 µm holes at 140 mm pitch

F. Sauli, Nucl. Instrum. Methods A386(1997)531

Thin, metal-coated polymer foil with high density of holes. On application of a voltage difference, each hole acts as an individual proportional counter, multiplying the electrons entering from the drift region. The amplified electrons leave the hole; most of the ions are collected by the upper electrode:

5-10,000 INDEPENDENT PROPORTIONAL COUNTERS per cm2

14


MULTIGEM DETECTORS MULTIGEM DETECTORS GEM electrodes can be cascaded, injecting the electrons released in one into the next multiplier; cascades of up to five GEMs have been tested. This allows to obtain larger gains, or safer operating conditions for the same gain. The track coordinates are obtained from the charge collected on strips or pads:

S. Bachmann et al, Nucl. Instr. and Meth. A479 (2002) 294

DRIFT

INDUCTION

GAIN

15


GEM PERFORMANCES GEM PERFORMANCES

65 µm rms

2-D POSITION ACCURACY:

M. Alfonsi et al, NIMA518(2004)106

LONG-TERM RADIATION RESISTANCE:

~ 4 1014 particles cm-2

3.106 particles mm-2

RATE CAPABILITY:

S. Bachmann et al, Nucl. Instr. and Meth. A479(2002)294

16


COMPASS TRIPLE GEM COMPASS TRIPLE GEM

Honeycomb plates

GEM foils

2-D Readout board

C. Altumbas et al, Nucl. Instrum. Methods A490(2002)177

Sturdy, light construction used for the tracker of the COMPASS experiment at CERN.30x30 cm2 active, 2-dimensional electronic readout 22 detectors are operational since 2002 in high intensity beam.

17



HALF-MOON TRIPLE GEM(TOTEM)40 detectors in installation at CERN LHC

Can be built in a variety of shapes, including non-planar

10-Chambers telescope:

18



CYLINDRICALPrototype for NA49 upgrade

Two sectored GEM foils, 60 cm long(Gas Detectors Development at CERN)

60 cm

19


INTERACTION VERTEX IMAGING (IVI) SETUPINTERACTION VERTEX IMAGING (IVI) SETUP

In collaboration with A. Scribano (Siena University)In collaboration with A. Scribano (Siena University)

Large angle Single-arm charged particle detectors:GEM detectors with fast electronic readout.Active during therapeutic exposures: the incoming beam has known position but cannot be detected

20


INTERACTION VERTEX IMAGING - SIMULATIONSINTERACTION VERTEX IMAGING - SIMULATIONS

Simulation of primary interactions for 400 MeV/u 12C beam show a large yield of charged prongs (mostly protons) emitted along the trajectory; this can be exploited for in-beam dose monitoring. There is however a large halo given by secondary vertices, mostly in the forward direction.

Primary Secondary

P. Solevi, Study of an in-beam PET system for CNAO. PhD Thesis at the University of Milano (2007)

Total

Secondary

The drop-off slope of the distribution provides information on the Bragg spectrum; it should be studied experimentally.

21


INTERACTION VERTEX IMAGING INTERACTION VERTEX IMAGING

Correlation plots between real and reconstructed vertex

EFFECT OF TRACK ENERGY AND ANGLE SELECTIONE>100 MeV, cosθ<0.9

Along the beam:

Perpendicular to the beam:

22


NUCLEAR SCATTERING TOMOGRAPHY (NST) NUCLEAR SCATTERING TOMOGRAPHY (NST)

At high proton energy (~600 MeV), recording of the protons scattered by the target and reconstruction of the interaction vertex provides the 3-Dimensional density distribution. From the subset of elastic proton-proton scattering, one gets the hydrogen density distribution.

Elastic:

The method was developed 25 years ago by G. Charpak and collaborators.

J.C. Duchazeaubeneix et al, J. Comp. Assisted Tomography 4 (1980)803

p

23


NUCLEAR SCATTERING TOMOGRAPHY NUCLEAR SCATTERING TOMOGRAPHY

Requires the development of very fast (> 1 MHz) data acquisition electronics.

The p-p cross section has a maximum at 1200 GeV/c (~ 600 MeV); about 50% of the interactions are elastic.Elastic events are selected from 2-tracks data using angular correlation and coplanarity.

Target volume of 10x10x10 cm3

10 mm3voxels --> 105 voxels.For 5% statistical accuracy on local density (500 events/voxel) ~5.104 events/cm3 Scattering probability ~ 10-3 cm-1

Total beam flux ~ 108 cm-2

Total number of events ~ 5.107

Total acquisition (at 1 MHz) ~ 1 minTotal dose ~ 30 mGy

24


IN-BEAM PET IN-BEAM PET

Computed distribution of + emitters212 MeV/u and 343 MeV/u 12C beam:

Imaging of co-linear gammas from positrons emitted by isotopes produced by the 12C -target interactions: 11C, 10C, 15O

Dual-head scanner:

P. Solevi, Study of an in-beam PET system for CNAO. PhD Thesis at the University of Milano (2007)

25


IN-BEAM PET (CRYSTALS)IN-BEAM PET (CRYSTALS)

PHOTONIS XP85013 Micro-Channel Photomultiplier (MCP)8x8 anode pads Signal risetime/width: 0.6/1.8 ns

Preliminary study of localization and Depth of Interaction determination using segmented crystals and a multi-anode photomultiplier

5 LYSO crystals 60x30x12 mm3

The center of gravity of the signal distribution provides the X-Y coordinates, the width the DOI

X

D

Y

26


IN-BEAM PET (CRYSTALS)IN-BEAM PET (CRYSTALS)

Width: Depth of interaction

Real

Measured

~ 1.5 mm rms

D:3 mm12 mm

25 mm

Center of Gravity: Localization in the plane of the MCP

MCP22Na

Coincidence

Preliminary measurements with a collimated 22Na source

With a simple algorithm, one can obtain a DOI determination corresponding to ~ 1/3 of the crystal thickness.

27


TIME OF FLIGHT PET (TOF) WITH SOLID STATE SENSORSTIME OF FLIGHT PET (TOF) WITH SOLID STATE SENSORS

Single photoelectron time resolution as a function of voltage:

G. Collazuol et al, Nucl. Instr. and Meth. A581(2007)461

Array of individual Avalanche Photodiodes, operated in the Geiger mode: Single photon counting, very good time resolution

SILICON PHOTOMULTIPLIER (SiPM)GEIGER AVALANCHE PHOTODIODE (G-APD) MULTI-PIXEL PHOTON COUNTERS (MPPC)

28


SiPMSiPM

Developed by various laboratories and commercially available

1x1 mm2

3x3 mm3

6x6 mm2 array

Possible detector assembly: array of independent fingers

Position-sensitive array:

MAJOR OPEN ISSUES:Light collection efficiency? Saturation? Intrinsic time resolution (crystal+sensor+electronics)? Radiation resistance? Cost?

Hamamatsu MPPC:

29


TOF PET: RESISTIVE PLATE CHAMBERSTOF PET: RESISTIVE PLATE CHAMBERSParallel plate gaseous avalanche chambers have excellent time resolution for charged particles. The use of resistive electrodes limits the dead time after a discharge:

100 µm GAP

ANODE:RESISTIVE GLASS

CATHODE

PICK-UP STRIPS

HIGH PRESSURE VESSEL

Yu.Pestov, Nucl. Instr. and Meth. 196(1982)45

The best resolution are obtained with high pressures (10 bar) and narrow gaps (100 µm):

Time resolution vs overvoltage for MIPs:

30


MULTIGAP RESISTIVE PLATE CHAMBER (RPC)MULTIGAP RESISTIVE PLATE CHAMBER (RPC)

E. Cerron Zeballos et al, Nucl. Instr. and Meth. A 374(1996)132

A stack of thin RPC gaps provides good efficiency and time resolution at atmospheric pressures; only one HV used, with floating middle plates. The signal is detected on external pickup electrodes.

10 gap, double stack module (ALICE) Efficiency and time resolution for fast charged particles:

~96% ~50 ps

31


MULTIGAP RPC TIME OF FLIGHT FOR ALICEMULTIGAP RPC TIME OF FLIGHT FOR ALICE

Adopted for the Time of Flight detector of ALICE: large area, very good time resolution (~50 ps rms)

RPC strip, ~120x12 cm2

~ 4 m

~ 7.5 m

The ALICE Experiment at CERN LHCJ. of Instrum. JINST 3 S08002 (2008)

32


IN-BEAM PET WITH RESISTIVE PLATE CHAMBERSIN-BEAM PET WITH RESISTIVE PLATE CHAMBERS

Efficiency for 511 keV as a function of layers (lead glass converters):

ADVANTAGES: TOF (~ 50 ps)• Large areas at low cost -> Full body PET • Arbitrary read-out pattern (strips, pads...) • DOI (thin modules)DRAWBACK:• No energy resolution

Multi-layer thin-gap RPC: the resistive electrode act as converters for photons:

M. Couceiro et al, Nucl. Instr. and Meth. A580(2007)915

Development at Coimbra University (P. Fonte)

33


AQUA RPC-PET DEVELOPMENT AQUA RPC-PET DEVELOPMENT

Development of a prototype multi-gap RPC with thin (300 µm) glass converter plates. To improve stability of operation, the glass is coated with a high-resistivity diamond-like layer. Thin nylon wires (fishing lines) maintain the gaps. The stack is filled with the operating gas mixture and sealed; the 2-D readout electrodes are external to the stack.

In collaboration with T. Tabarelli (Univ. La Bicocca, Milano).

OPEN QUESTIONS:• Efficiency: choice of best electrodes (must be dielectric)• Time and space resolution• Tolerance to other radiations (neutrons, protons)• Gas choice, Long-term behaviour ......

34


RATES AND EXPOSURE TIME RATES AND EXPOSURE TIME

Therapeutic beam intensity ~ 1010 p/s on ~1cm2

Raster beam scan: partly overlapping spots moved every 5 ms (~5.107 p/spot)

PRESENT DETECTORS LIMITATION: GEM Tracker: 107 p/cm2 (intrinsic due to space charge) Multi-events resolving time (charge collection+electronics shaping) ~100 ns Total flux over 10x10 cm2 for 10% pileup ~ 106 p/s ---> 1 MHz Electronics readout ~ 100 kHz (105 p/s) RANGE TELESCOPE: Intrinsic (scintillator+WLS decay time) ~ 30 ns Silicon PM+shaping ~ 100 ns --> 10 MHz Electronics readout ~ 100 kHz

PRESENT PERFORMANCES PRR: Pixel size 10 mm2 --> 103 pixels for 10x10 cm2 image For 3% statistical error (103 events/pixel) ~ 106 tracks --> 10 s for 100 kHz readout NST: Voxel size 5x5x5 mm3 --> 2.5 104 voxels for 10x10x30 cm3 target For 3% statistical error (103 events/voxel) --> 2.5 107 events --> 250 s (~4 min) for 100 kHz readout

Estimated do not include efficiency losses, data analysis time etc.

beam intensity should be reduced by four orders

data acquisition rate could be increased 10x

BETTER RESOLUTION-SHORTER ACQUISITION TIMES: 100 kHz--> >1 MHz

35


Date post:	23-Dec-2015
Category:	Documents
Upload:	terence-bates
View:	212 times
Download:	0 times

F. Sauli - ICPT Workshop (Oxford, 23-26 September 2008) 0 AQUA A DVANCED QU ALITY A SSURANCE FOR...

Documents