Hadron Imaging and X-ray dosimetry systems in clinical radiotherapy
Mara BruzziUniversity of Florence, Dept. of Physics and Astronomy and INFN Firenze, Italy
M. Bruzzi, Scuola di Specializzazione in Fisica Medica, Milano , May 10, 20191
Outline
M. Bruzzi, Scuola di Specializzazione in Fisica Medica, Milano , May 10, 2019
1. Motivation
2. Radiotherapy Irradiation Facilities
3. X-ray dosimetry systems in clinical radiotherapy
❑ epitaxial silicon
❑ polycrystalline diamond
4. Proton Imaging in Proton Therapy
❑ proton Computed Tomography – pCT
5. Conclusions
2
RadiotherapyA technological solution to a biological problem
Radioterapy with external beamsRadiation beams with high energy (X, γ,electrons, protons, ions) produced byradionuclides or particle accelerators
BrachytherapySealed radioactive sources introduced in the body
Methabolic radiotherapyNon-sealed radioactive sources vehicolatedwithin the body
3M. Bruzzi, Scuola di Specializzazione in Fisica Medica, Milano , May 10, 2019
•immobilization techniques allow forpositioning patient during the treatment withalways increasing accuracy,
•accelerators equipped with beam modifyingdevices able to get dose distributions closelyshaped on the target volume.
State-of-art Radiotherapy Machines
Procedures have been consistently upgraded in recent years, dueto continuously developing technologies.
Advanced, highly accurate imaging and dose verifications systems must match with increased accuracy of irradiation techniques .
4M. Bruzzi, Scuola di Specializzazione in Fisica Medica, Milano , May 10, 2019
M. Bruzzi, Physics and Astronomy Dept. Univ. of Florence– Physicist, materials science and detector development
C. Civinini, Physicist, INFN Sezione di Firenze- Physicist, pCT tracker development
M. Scaringella, INFN Florence Tecnologo- Electronic Engineer
M. Intravaia, PhD Student, Pegaso Univ. of Pisa/Florence/Siena- Electronic Engineer
C. Talamonti, Biomedical, Experimental and Clinical Science Dept. Univ. of Florence– Medical Physicists
The Florence Research Group
5
LINAC radiotherapy facility• LINAC ( Linear Accelerator ) by means of high
frequency ( ~ 3 GHz ) electromagnetic wavesaccelerates charged particles at high energy along a linear path
M. Bruzzi, Scuola di Specializzazione in Fisica Medica, Milano , May 10, 2019
X-raysirradiation
6M. Bruzzi, Scuola di Specializzazione in Fisica Medica, Milano , May 10, 2019
M. Bruzzi, Scuola di Specializzazione in Fisica Medica, Milano , May 10, 2019
Intensity Modulated Radiation Therapy (IMRT) consists in using a fewradiation beams, generally from 2 to 9, produced by the same linearaccelerator and directed towards the tumor from different angles, in order toconcentrate the dose released on the volume of the tumor.
Intensity Modulated Radiation Therapy (IMRT)
7
M. Bruzzi, Scuola di Specializzazione in Fisica Medica, Milano , May 10, 2019
Flat Dose Map Modulated Intensity Dose Map
To spare at best surrounding healthy tissues the dose released to the tumor needs to be shaped along an
irregular field to be best conformed to the tumorvolume
8
Multi Leaf Collimators (MLC)The dose conformation is obtainedusing Multileaf Collimators with setsof mobile lamellas in W mountedexternally on the LINAC head
9
M. Bruzzi, Scuola di Specializzazione in Fisica Medica, Milano , May 10, 2019
M. Bruzzi, Scuola di Specializzazione in Fisica Medica, Milano , May 10, 2019
Example: IMRT irradiation in step and shoot modality, the total dose is released asa sum of nine segments. Each segment corresponds to a particular arrangement ofthe MLC lamellas (left). Beam is off during their movement.
10
Total dose is obtained as the sum of the dose released by each segment (right).
M. Bruzzi, Scuola di Specializzazione in Fisica Medica, Milano , May 10, 2019
VMAT (Volumetric Modulated Arc Therapy)
❑ Able to focus more accurately at tumor tissues, ensuring greater preservation ofhealthy ones.
❑ Modulating not only the amplitude and velocity of the MLC, but also rotationspeed of the Gantry and Linac dose-rate.
❑ continuous rotation of the accelerator head during irradiation for maximumfocusing of radiation on tumor tissues, which are thus affected by all possibleangles.
❑ significantly reducing duration of treatments compared to IMRT: about 5-7minutes compared to traditional times which are around 20 minutes persession.
❑ useful when treatment focus must be maximum to preserve nearby organs:tumors of the head / neck, as larynx, pharynx and oral cavity; tumors of thepelvis, as prostate and rectum; tumors of the lung and breast.
11
Hadrontherapy
Hadrontherapy is the treatment of tumors by heavy charged particles, such
as protons (proton-therapy) or ions.
Sviluppo di un sistema di imaging
tomografico con protoni per adroterapia
Firenze, 4 Ottobre 2018
2
➢ X-rays, often used in clinical radiotherapy, release their energy gradually during
their propagation.
➢ Protons release most of their energy in the Bragg peak.
Accurate dose distribution planning
12M. Bruzzi, Scuola di Specializzazione in Fisica Medica, Milano , May 10, 2019
Outline
M. Bruzzi, Scuola di Specializzazione in Fisica Medica, Milano , May 10, 2019
1. Motivation
2. Radiotherapy Irradiation Facilities
3. X-ray dosimetry systems in clinical radiotherapy
❑ epitaxial silicon
❑ polycrystalline diamond
4. Proton Imaging in Proton Therapy
❑ proton Computed Tomography – pCT
5. Conclusions
13
M. Bruzzi, Scuola di Specializzazione in Fisica Medica, Milano , May 10, 2019
Highly accurate imaging and dose verifications systems must match with
increased accuracy of the irradiation techniques . In IMRT / VMAT, the way to
perform dose distribution verifications is still a matter of discussion within the
scientific community. Detector requirements:
• Response independent of energy - often Charged Particle Equilibrium (CPE)
lacks (a phenomenon associated with the range of secondary particles and hence
dependent on the beam energy, composition and density of the medium );
• small volume and high sensitivity - to get enough spatial resolution;
• response independent of dose rate, continuously changing during VMAT;
• Real time invivo detectors - european community require dose delivery to be
verified experimentally directly during irradiation (Article 56 of COUNCIL-
DIRECTIVE-2013/59/EURATOM).
Dosimetry Challenges
14
M. Bruzzi, Scuola di Specializzazione in Fisica Medica, Milano , May 10, 2019
State-of-art commercial dosimetric devices used in clinicalradiotherapy
IC (AIR) SILICON DIAMOND
[g/cm3] 1.29x10-3 2.33 3.52
Ei [eV] 34.00 3.60 16.20
S [nC/Gymm3] 0.038 647.22 217.28
Area [mm2] 25.00 0.64 3.80
thickness [mm] 5.00 0.03 0.001
volume [mm3] 125 0.019 0.0038
Array OCTAVIUS
PTW
MAPCHECK
SunNuclear
-
Detector 729 PTW SunPoint®
Diode Detector
microDiamond
type 60019 PTW
Reference
15
M. Bruzzi, Scuola di Specializzazione in Fisica Medica, Milano , May 10, 2019
16
M. Bruzzi, Scuola di Specializzazione in Fisica Medica, Milano , May 10, 201917
Advantages:- High sensitivity (about 18000 times higher than air filled IC with same activevolume).- Well developed manufacture technology.- high spatial resolution.- work in null bias mode ( in-vivo applications possible ).
Drawbacks:- Sensitivity decrease with accumulated dose due to increase of concentrationof recombination centers (recalibrations needed).- Dose rate dependency due to centers saturation at high dose rates.- Energy dependence, since Si is not "water equivalent" (Z=14).
The Silicon Choice
M. Bruzzi, Scuola di Specializzazione in Fisica Medica, Milano , May 10, 2019
18
19
iSi ERG /=
G,: carrier generation rate:
R,: dose rate;
Si: Si density;
Ei: mean ionization energy.
s: sensitivity
(per unit volume of active region)
Q: released charge;
D: delivered dose;
q: elementary charge
3637
mmGy
nC
E
q
R
qG
D
Qs
i
Si
====
Ei vs Egap constant for many
semiconductors under general
conditions. In silicon
Ei≈3.6eV/pair.
a) "null bias" (to minimize leackage current).
b) DC coupling.
c) Sampling time and reset fixed by digital electronics
(usually T10ms).
d) Only integrated charge is measured.
Working Principle
0 2 4 6 8 10 12 140
100
200
300
400
500
n-type
p-type
Sensitiv
ity [nC
/Gy]
Dose [KGy]
G.Rikner et al.Phys. Med. Biol. 28, 1983, 1261-1267
Decrease in sensitivity S during device lifetime in past-times
- Frequent Calibration needed. - Usually Si devices pre-irradiated up to 10kGy
M. Bruzzi, Scuola di Specializzazione in Fisica Medica, Milano , May 10, 2019
20
The Florence Radiation Hardening Choice: Epitaxial Si
• 4” p-type MCz wafer
• Epitaxial layers with 50mm thickness grown by ITME Warsaw Poland
• Manufacturing of the device carried out by ITC-IRST Trento (now FBK)
Constant active volume during irradiation: epitaxial layer thickness andguard ring distance from active area thinner than minimum diffusion lengthreached during device lifetime will provide constant sensitivity.
High quality crystalline epitaxial Si of 50mm thickness is now commerciallyavailable on large scale at very low costs
Sensitivity of silicon is quite high, and the subsequent reduction in signalstrength is of no concern.
M. Bruzzi, Scuola di Specializzazione in Fisica Medica, Milano , May 10, 2019
21
Radiation hardening: thin p-type Epitaxial Si
M. Bruzzi, Scuola di Specializzazione in Fisica Medica, Milano , May 10, 2019
22
Pixel: 2x2mmPitch: 3mmDetector size: 6.3x6.3cm2
Matrix: 21x21pixels
M. Bruzzi, Scuola di Specializzazione in Fisica Medica, Milano , May 10, 2019 23
Nine modules to cover an area of about 2020 cm2. System complexity: ~4k channels
2D Dosimeter Module and Array Geometry
24
The Diamond Choice
◼ it is almost water equivalentit doesn’t perturb the radiation field → small fields the energy is absorbed as in the water → no correction factors
◼ high radiation hardness → long term stability◼ high density → high sensitivity → small dimensions◼ non toxic◼ it can be used as TL dosimeter (off-line) or for on-line applications
☺
◼ high defect density - priming effects – instability of the signal◼ high voltage required◼ high production costs
M. Bruzzi, Scuola di Specializzazione in Fisica Medica, Milano , May 10, 2019
No large area array available yet
M. Bruzzi, Scuola di Specializzazione in Fisica Medica, Milano , May 10, 2019
25
26
Single crystal diamond dosimeter
Layout device produced by Università di Roma Tor Vergata
26M. Bruzzi, Scuola di Specializzazione in Fisica Medica, Milano , May 10, 2019
The Florence Choice: polycrystalline Diamond
◼ difficult to selectstones with proper
dosimetric response
Natural diamond
Single crystal CVD (Chemically VapourDeposited) diamond
Polycrystalline CVD diamond
◼ zero/low voltage to reduce polarization effects*
◼ grown on HPHT diamond◼ not available in large areas
❑ ability to produce largearea wafers of 3-5”
☺
*M. Bruzzi et al., Diamond & Related Materials 20 (2011) 84–92
M. Bruzzi, Scuola di Specializzazione in Fisica Medica, Milano , May 10, 2019
27
28
Chemical Vapour Deposited polycrystalline Diamond
50 mm 200mm Courtesy of Element Six
After polishing andMaterial removal
- Columnar growth –increased quality at
growth side
M. Bruzzi, Scuola di Specializzazione in Fisica Medica, Milano , May 10, 2019
28
Problem: priming and instability effects due to defects in pCVD
M. Bruzzi, Scuola di Specializzazione in Fisica Medica, Milano , May 10, 2019
29
M. Bruzzi, Scuola di Specializzazione in Fisica Medica, Milano , May 10, 2019
TO GET STABLE SIGNAL
Idea : back to back schottky barriersand low/null bias
defects in the bulk are notcontributing to the currentresponse and signal is thus onlydue to the charge collected atinterfaces.
Trapping–detrapping mechanismsin the bulk, which are believed togive the main contribution topriming and signal instability aretherefore now negligible.
30
The Polycrystalline pixel diamond dosimeter prototype of Florence
b)
- Material- Up to three polycrystalline diamond films 2.5x2.5cm2 active area each, 300mm thick; - Premium Detector Grade Element Six, UK
- Contacts • Schottky Barriers produced @ University of Florence•12 x 12 matrix, pixel size: 1.8x1.8 mm2 → 288 pixels in total
- Read Out Electronics •four 64 channels 20 bit current-input analog to digital converter chips able of measuring currents from fAs to mAs; 160ms-1s integration time (50ms)•custom printed circuit board;•semi-rigid silver-polymer pin-contacts produced by us connecting each pixel of the 144 matrix connecting vias on PCB .
-Measurement•Low voltage to get fast and reproducible signals;•Device can be moved in x-y directions to cover a wider radiation field area.
M. Bruzzi, Scuola di Specializzazione in Fisica Medica, Milano , May 10, 2019
31
pCVD Diamond test under linac
LINAC @ Radioterapia AOUC Firenze
1 pCVD on PCB 2 pCVD in PMMA /front
/rear
M. Bruzzi, Scuola di Specializzazione in Fisica Medica, Milano , May 10, 2019
32
Linearity with dose
14th IPRD, 3 - 6 October 2016 Siena, Italy A. Bartoli et al.
33M. Bruzzi, Scuola di Specializzazione in Fisica Medica, Milano , May 10, 2019
Repeatability
Dose rate dependence
∆=0.9832 ± 0.0015
𝑆 =𝑑𝑄
𝑑𝐷~24
𝑛𝐶
𝐺𝑦
Performance under conventional and IMRT radiotherapy beam
Current response of all pixels in a
conventional X-ray beam (Vapp = 1V)
Dose-rate 50 Mu/min
✓negligible dark current → high S/N ✓negligible polarization effects → stable
response , fast dynamics
Current response of one pixel under
an IMRT beam in step and shoot
modality
Bartoli et al. 2017 JINST 12 C03052
M. Bruzzi, Scuola di Specializzazione in Fisica Medica, Milano , May 10, 2019
34
2.5x2.5cm2 pCVD Diamond prototype IMRT map 14x10cm2
measured by shifting the diamond dosimeter
IMRT breast cancer map asmeasured by the pCVD Diamond .
IMRT breast cancer map as calculated by the TPS (treatment planning system)
(GT = gantry target direction; LL = lateral-lateral direction) Grid spacing 3 mm.
First IMRT map with Diamond Device
M. Bruzzi, Scuola di Specializzazione in Fisica Medica, Milano , May 10, 2019
35
Time structure of VMAT
delivery as measured by
one pixel of the pCVD
diamond dosimeter
VMAT Experimental Test
- lung cancer treatment fraction- 2 polycrystalline diamond dosimeters - Active area: 5.0x2.5cm2
- moved ±2.0cm in the y-direction to cover field
M. Bruzzi, Scuola di Specializzazione in Fisica Medica, Milano , May 10, 2019
36
Calculated vs measured VMAT maps
➢ VMAT lung cancer treatment ➢ 2 polycrystalline diamond dosimeter 2.5x2.5cm2 moved in xy directions to cover 6x5cm2 area
VMAT map as calculated by the TPS (Treatment Planning System)
VMAT map as measured by thepCVD Diamond .
90% pixels with Y index< 1
(3mm, 3% criteria)
Bartoli et al. 2017 JINST 12 C0305237
M. Bruzzi, Scuola di Specializzazione in Fisica Medica, Milano , May 10, 2019
M. Bruzzi, Scuola di Specializzazione in Fisica Medica, Milano , May 10, 2019
What Next ? The Florence Choice: CsPbBr3
38
M. Bruzzi, Scuola di Specializzazione in Fisica Medica, Milano , May 10, 2019
Present project in Florence aims at developing a new dosimetric system for in vivo dosimetry, exploiting flexible geometry, real-time measurement, wireless read-out.
microcrystalline CsPbBr3 film deposited on alumina substrate carrying two parallel gold contacts
In CsPbBr3, mean energy to create an electron-hole pair: ECsPbBr3 = 5.3 eV. Considering CsPbBr3 = 4.55 g/cm3 → sCsPbBr3 = 860 nC/ Gymm3
Higher sensitivity than silicon:
TPS treatment plan Relative dose verification TPS treatment plan, Si Mapcheck and CsPbBr3.
39
Outline
M. Bruzzi, Scuola di Specializzazione in Fisica Medica, Milano , May 10, 2019
1. Motivation
2. Radiotherapy Irradiation Facilities
3. X-ray dosimetry systems in clinical radiotherapy
❑ epitaxial silicon
❑ polycrystalline diamond
4. Proton Imaging in Proton Therapy
❑ proton Computed Tomography – pCT
5. Conclusions
40
Proton Therapy Challenges: proton Computed Tomography - pCT
A treatment plan needs the patient’s Relative (to water)Stopping Power (RSP) maps.
RSP maps are currently extracted from x-ray CTs, introducingerrors in the Bragg peak positioning up to a few millimetres.
The direct measurement of the proton RSP maps, usingprotons themselves for tomographies, can potentially reduceinaccuracies in tumor irradiation.B. Schaffner and E. Pedroni Phys. Med. Biol. 43 (1998) 1579–1592
41
𝑆(𝐸)𝑅 =𝑆(𝐸)
𝑆(𝐸)𝑤𝑎𝑡𝑒𝑟𝑆 𝐸 = −
𝑑𝐸
𝑑𝑙
M. Bruzzi, Scuola di Specializzazione in Fisica Medica, Milano , May 10, 2019
A proton CT apparatus
Prima – RDH – IRPT Collaboration
M. Bruzzi1,2, C. Civinini2, M.Intravaia3,2, N. Randazzo4, M. Rovituso4,
M. Scaringella2, V. Sipala5,6, F. Tommasino4,7
1Physics and Astronomy Department, University of Florence, Florence, Italy 2INFN - Florence, Florence, Italy
3Information Engineering and Mathematical Sciences Department, University of Siena, Italy4INFN - Catania, Catania, Italy
5INFN - TIFPA, Trento, Italy6INFN - Laboratori Nazionali del Sud, Catania, Italy7Chemistry and Pharmacy Department, University of Sassari, Sassari, Italy8Physics Department, University of Trento, Trento, Italy
M. Bruzzi, Scuola di Specializzazione in Fisica Medica, Milano , May 10, 2019 42
The pCT system➢ Measurement of the trajectory of
each proton. Tracker
➢ Measurement of the residual energy
of each proton.Calorimeter
➢ Most Likely Path reconstruction.
➢ Tomographic reconstruction.
M. Bruzzi, Scuola di Specializzazione in Fisica Medica, Milano , May 10, 2019
4
Firenze, 4 Ottobre 2018
3x3x10 cm^3
YAG:Ce
crystals
43
Tracking with multiple scattering
Actual proton trajectory
L → curved trajectory with narrow confidential limits→ Most Likely Path of the proton inside the object.
Measurements: entry andexit positions and angles
LMost Likely Path (MLP)calculation
+
M. Bruzzi, Scuola di Specializzazione in Fisica Medica, Milano , May 10, 2019
44
INFN-Prima pCT apparatus
Beam test at Trento proton Therapy Centre experimental beam lineProton energy: nominal 211 MeV (198 MeV at phantom) 5x20 cm2 field-of-view
M. Bruzzi, Scuola di Specializzazione in Fisica Medica, Milano , May 10, 2019
45
Tracker architecture
X-side view1° level FPGA
2° level FPGA
M. Bruzzi, Scuola di Specializzazione in Fisica Medica, Milano , May 10, 2019
46
Sviluppo di un sistema di imaging
tomografico con protoni per adroterapia
Firenze, 4 Ottobre 2018
10
Electron Density Phantom
used for calibration Anthropomorphous phantom Radiography showing
implanted metallic
prosthesys
Phantoms
Phantoms are used during tests on proton beam to simulate the presence
of the patient and to determine the ultimate performance of the apparatus.
CIRS Proton Therapy dosimetry head. Mod. 731 HN
47M. Bruzzi, Scuola di Specializzazione in Fisica Medica, Milano , May 10, 2019
Anthropomorphous phantomX-Ray radiographies
Tomography region
M. Bruzzi, Scuola di Specializzazione in Fisica Medica, Milano , May 10, 2019
48
Anthropomorphous phantom tomography
Matteo Intravaia - INFN Firenze 14th Trento Workshop - Trento
February 27th 2019 49
Anthropomorphous phantom tomography
64 axial slicesvoxels: 600x600x812mm3 0.3mm3
400 angles (0.9 deg. uniform spacing)About 3.7 x 107 events (selected)
First and last slices have low statistics (outside field of view)Movie starts from lower jaw ends at upper teeth (5.2 cm range)
Total estimated dose 1.5mGy
Much lower contrast with respect to the X-Ray images mainly because of the physics of the interactions: important Z dependence for X-Ray, density dependence for protons→BUT it is what is needed for hadron therapy.
M. Bruzzi, Scuola di Specializzazione in Fisica Medica, Milano , May 10, 2019
50
1.2x108 events in 400 angles
Metallic prosthesis artifacts• In presence of implanted metal prosthesis, the treatment plan is difficult
to define because of xCT severe artifacts, which degrade the image quality and negatively influence the RSP map determination.
• pCT is less sensitive to high Z materials than xCT→more accurate RSP maps nearby metallic implants.
Tests are planned to quantify the influence of these structures on the treatment quality
pCT xCT
M. Bruzzi, Scuola di Specializzazione in Fisica Medica, Milano , May 10, 2019
51
Electron density phantom→8 lateral plugs:1) Liver 1.07 gcm-3
2) Lung exhale 0.50 gcm-3
3) Breast 0.99 gcm-3
4) Bone 1.53 gcm-3
5) Muscle 1.06 gcm-3
6) Bone 1.16 gcm-3
7) Adipose 0.96 gcm-3
8) Lung inhale 0.20 gcm-3
→Central plug:Liquid water vial
→‘Water equivalent’ plastic material (Plastic Water LR) bulk.
1
2
3
4
5
6
7
8
M. Bruzzi, Scuola di Specializzazione in Fisica Medica, Milano , May 10, 2019
52
Electron density phantom tomography
• 1.6 mm thick central horizontal slice
• All inserts are visible
• Quantitative analysis of measured vs expected RSP correlation (next slide)
Next slide section
M. Bruzzi, Scuola di Specializzazione in Fisica Medica, Milano , May 10, 2019
53
108 events in 400 angles
RSP correlation
Excellent correlation (deviation < 1%)!
M. Bruzzi, Scuola di Specializzazione in Fisica Medica, Milano , May 10, 2019
54
Conclusions• Advanced imaging and dosimetry systems needed to
get increased accuracy in radiotherapy treatment plans;
• Diamond potentially best suited for X-ray radiotherapy due to energy independence of its response (tissue equivalence );
• Diamond dosimeter array not yet available commercially; a prototype manufactured and tested in Florence under IMRT and VMAT treatment plans shows promising results;
M. Bruzzi, Scuola di Specializzazione in Fisica Medica, Milano , May 10, 2019
• Increased accuracy in proton therapy by direct evaluation of Relative Stopping Powers
• proton Computed Tomography (pCT) system manufactured and testedunder a 200MeV p beam for pre-clinical study at Trento proton Therapy Centre;• Measured RSP values agree with the expected ones at level of 1%; • Technique proves to be potentially beneficial in the presence on prosthesys
55