Mitglied der Helmholtz- Gemeinschaft Designing and Improving Medical Imaging Systems by Monte Carlo Studies using GATE Institute of Neurosciences and Medicine (INM) Research Center Juelich & Department of Mathematics and Natural Sciences University of Wuppertal Germany Uwe Pietrzyk
58
Embed
Designing and Improving Medical Imaging Systems by Monte Carlo Studies using GATE
Designing and Improving Medical Imaging Systems by Monte Carlo Studies using GATE. Institute of Neurosciences and Medicine (INM) Research Center Juelich & Department of Mathematics and Natural Sciences University of Wuppertal Germany Uwe Pietrzyk. Outline: - PowerPoint PPT Presentation
Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Mitg
lied
der H
elm
holtz
-Gem
eins
chaf
t
Designing and Improving Medical Imaging Systemsby Monte Carlo Studies using GATE
Institute of Neurosciences and Medicine (INM)Research Center Juelich
&Department of Mathematics and Natural Sciences
University of WuppertalGermany
Uwe Pietrzyk
ICATPP October 2009 U Pietrzyk - INM – FZ-Juelich - Germany Slide 2
Outline:
• Intro: Neuroscience in the Research Center Jülich
• Motivation: Why using Monte-Carlo Techniquesfor Simulation?
• Some Basics on Medical Imaging
• The Concept of GATE / GEANT4 / ROOT
• Examples of GATE Applications
• Summary & Demo
ICATPP October 2009 U Pietrzyk - INM – FZ-Juelich - Germany Slide 3
Outline:
• Intro: Neuroscience in the Research Center Jülich
• Motivation: Why using Monte-Carlo Techniquesfor Simulation?
• Some Basics on Medical Imaging
• The Concept of GATE / GEANT4 / ROOT
• Examples of GATE Applications
• Summary & Demo
ICATPP October 2009 U Pietrzyk - INM – FZ-Juelich - Germany Slide 4
Institute of Neurosciences and MedicineArea for 9.4T MR/PET
ICATPP October 2009 U Pietrzyk - INM – FZ-Juelich - Germany Slide 5
Moving the Magnet for the 9.4T MR/PET Hybrid Scanner
ICATPP October 2009 U Pietrzyk - INM – FZ-Juelich - Germany Slide 6
9.4T MR/PET Hybrid Scanner / Magnet in place / PET-Module on site
ICATPP October 2009 U Pietrzyk - INM – FZ-Juelich - Germany Slide 7
Institute of Neurosciences and Medicine (INM)
Architectonics and brain functionProf. Dr. K. Amunts
Structural und Functional
Organisation of the Brain
Prof. Dr. K. Amunts
Cognitive Neurology
Prof. Dr. G.R. Fink
Physics of MedicalImaging
Prof. Dr. N.J. Shah
MR PhysicsProf. Dr. N.J. Shah
PETProf. Dr. H. Herzog
Brain tumoursProf. Dr. K.-J. Langen
Molecular Organisation of the Brain
Prof. Dr. K. Zilles
Transmitters-receptors
Prof. Dr. K. Zilles
Structureof Synapses
Prof. Dr. J. Lübke
Molecular Neuroimaging
Prof. Dr. A. Bauer
Functional neuronal circuits
Prof. Dr. D. Feldmeyer
System MedicineProf. Dr. Dr. P. A. Tass
NeurotechnologyPD Dr. C. Hauptmann
Neuromodulation
Prof. Dr. Dr. P. A. Tass
MathematicalNeuroscience
PD Dr. O.V. Popovych
RadionucleiDevelopmentDr. B. Scholten
Radio-pharmacology
Dr. D. Bier
RadiotracerDevelopment
Dr. D. Holschbach
RadiotracerProductionDr. J. Ermert
Dr. K. Hamacher
Nuclear Chemistry
Prof. Dr. H.H. Coenen
Ethics in theNeurosciences
Prof. Dr. D. Surma
Systems Biologyand
Neuroinformatics
N.N.
Multimodal ImageProcessing & Morphometry
Prof. Dr. U. Pietrzyk
ICATPP October 2009 U Pietrzyk - INM – FZ-Juelich - Germany Slide 8
Analysis of the structure and the functional processes of the brain at the organ level and at the cellular level. To understand the organisational principles of the brain.
To explore the mechanisms of the normal and pathological nervous system.
Development of new techniques for diagnosis and therapy for neurological and psychiatric diseases, e.g. demand-driven, deep brain stimulation, electrical or chemical neuromodulation.
Development of new methodologies for imaging the in vivo brain.
ICATPP October 2009 U Pietrzyk - INM – FZ-Juelich - Germany Slide 9
Highlights from the Research Programme
Molecular PET Imaging
Mechanisms of Cognitive Processes in Normals and Patients
Brain Pacemaker
3D Map of the Human BrainNew MR Methods
Neurodegenerative Diseases
ICATPP October 2009 U Pietrzyk - INM – FZ-Juelich - Germany Slide 10
ICATPP October 2009 U Pietrzyk - INM – FZ-Juelich - Germany Slide 11
Outline:
• Intro: Neuroscience in the Research Center Jülich
• Motivation: Why using Monte-Carlo Techniquesfor Simulation?
• Some Basics on Medical Imaging
• The Concept of GATE / GEANT4 / ROOT
• Examples of GATE Applications
• Summary & Demo
ICATPP October 2009 U Pietrzyk - INM – FZ-Juelich - Germany Slide 12
Interesting for studying the features of detection systems, which cannot be described analytically:
• How to estimate the acceptance of a multichannel detector for high energy photons?
• How to determine the contribution from scattered photons prior to calculate the tracer uptake in a certain Region of Interest (ROI) image analysis?
Motivation (1):
Why using MonteCarlo techniques to simulate Medical Imaging Devices?
ICATPP October 2009 U Pietrzyk - INM – FZ-Juelich - Germany Slide 13
Advantage:
Can describe very complex systems!!
Disadvantage:
Often requires considerable computation times!!
It is a statistically founded method, hence, bearsintrinsical errors!! Samples have to be sufficiently large!!
Motivation (2): MonteCarlo techniques – Advantages and Disadvantages!
ICATPP October 2009 U Pietrzyk - INM – FZ-Juelich - Germany Slide 14
Processing Pipeline:
From basic detector design
to reconstructed imagesandquantitative evaluation
Note: Such images are reliable only, if we can handleall corrections. Simulation is an essential support!
Motivation (3): MonteCarlo technique – an interesting option!
ICATPP October 2009 U Pietrzyk - INM – FZ-Juelich - Germany Slide 15
Outline:
• Intro: Neuroscience in the Research Center Jülich
• Motivation: Why using Monte-Carlo Techniquesfor Simulation?
• Some Basics on Medical Imaging
• The Concept of GATE / GEANT4 / ROOT
• Examples of GATE Applications
• Summary & Demo
ICATPP October 2009 U Pietrzyk - INM – FZ-Juelich - Germany Slide 16
• The fundamental experimental setup
the main components
• Difference of functional and structural imaging:
PET and SPECT vs. CT and MRINuclear Medicine: Radiology:PET = Positron Emission Tomography CT= X-Ray Computed TomographySPECT= Single Photon Emission MRI= Magnetic Resonance Imaging Computed Tomography
Note: PET and SPECT are “counting experiments”!!
Some Basics on Medical Imaging
ICATPP October 2009 U Pietrzyk - INM – FZ-Juelich - Germany Slide 17
The Basic Principle of Imaging
Source (external)
Object (+ Source)
Selection / Definition using:
(a) Diaphragm; (b) Grid;
(c) Collimator;(d) Coincidence Circuit)
Detector
X
✓
ICATPP October 2009 U Pietrzyk - INM – FZ-Juelich - Germany Slide 18
Current Scene of Functional andStructural / Morphological Imaging
18
PET PET/CT CT
SPECT/CT
SPECTMRT
MR/PET
ImageFusion
ICATPP October 2009 U Pietrzyk - INM – FZ-Juelich - Germany Slide 19
Complementary Nature: example: MRI & PET
19
Note: Already today, PET is mostly available as a combined modality, namely PET/CT
ICATPP October 2009 U Pietrzyk - INM – FZ-Juelich - Germany Slide 20
Basics in Positron-Emission-Tomography (I)
20
Note Two co-linear photons No collimation!Need correction for scatter and attenuation!
Unknown tracer-distribution in an environment of unknown denstity
MRI
Detector Detector
ICATPP October 2009 U Pietrzyk - INM – FZ-Juelich - Germany Slide 21
Basics in Positron-Emission-Tomography (II)
212121
MRI
+ Imaging System: Detector with high resolution and high sensitivity
Scintillators (LSO / GSO, ...) coupled to PMT or APD fast electronics
+ highly specific tracers, „smart probes“; nano molar concentrations
ICATPP October 2009 U Pietrzyk - INM – FZ-Juelich - Germany Slide 22
Outline:
• Intro: Neuroscience in the Research Center Jülich
• Motivation: Why using Monte-Carlo Techniquesfor Simulation?
• Some Basics on Medical Imaging
• The Concept of GATE / GEANT4 / ROOT
• Examples of GATE Applications
• Summary & Demo
ICATPP October 2009 U Pietrzyk - INM – FZ-Juelich - Germany Slide 23
The Goals of Simulation in Nuclear Medicine:
GGeant4 AApplication forTTomographic EEmission
Scanner Design
Protocol Optimization
Image Reconstruction
Data Analysis
Testing new algorithms
Scatter Correction
Quantification / Recovery
ICATPP October 2009 U Pietrzyk - INM – FZ-Juelich - Germany Slide 24
Two different Approaches:
General purpose simulation codes (GEANT4, EGS4, MCNP…) wide range of physics wide community of developers and users documentation, maintenance and support complexity speed
Dedicated simulation codes (PETsim, SimSET, Eidolon, SIMIND…) optimized for nuclear medical imaging applications (geometry, physics...) ease of use and fast development maintenance, upgrades
ICATPP October 2009 U Pietrzyk - INM – FZ-Juelich - Germany Slide 25
A Combined Approach (I):
25
• Realistic modeling of PET/SPECT experiments modeling of detectors, sources, patient movement (detector, patient) time-dependent processes (radioactive decay,
movement management, biological kinetics) • Ease-of-use• Fast• Long-term availability, support and training
PET/SPECT dedicated
developmentsGATEGATE
(by OpenGATE Collaboration)
(by GEANT4 Collaboration)
ICATPP October 2009 U Pietrzyk - INM – FZ-Juelich - Germany Slide 26
A Combined Approach (2):
26
• Based on GEANT4object Oriented Analysis & Designwide range of physics modelslong term availabilityupgrades, documentation & support
• Specific developments regarding to Nuclear medical imaging needs
material database, sources, readouttime and movement management
• Ease-of-use for non C++ programmersscripting commands to define all paramaters of the simulation (construction of the geometry, specification of the physical processes involved, of the sources...)
ICATPP October 2009 U Pietrzyk - INM – FZ-Juelich - Germany Slide 27
GATE structure (1):
General Scope
27
• Three different levels:• GEANT4 core
• Developer level framework and application
classes C++ programming
• User level sequence of scripting
commandsgeometry construction physical processes involvedsources (geometry, activity) movement (type, speed…) duration of the acquisition
User interface
Application classes
Framework
Geant4
ICATPP October 2009 U Pietrzyk - INM – FZ-Juelich - Germany Slide 28
GATE structure (2):
Geometry Construction A specific mechanism has been
developed to help the user construct easily a geometry scripting commands geometry = combination of
geometric volumes, like ‘russian dolls’
ICATPP October 2009 U Pietrzyk - INM – FZ-Juelich - Germany Slide 29
GATE structure (2):
Geometry Constructionworld
Source
Body Head
Scanner
Rsector
Crystal
LSO BGO
ICATPP October 2009 U Pietrzyk - INM – FZ-Juelich - Germany Slide 30
GATE structure (2):
Geometry Constructionworld
Source
Body Head
Scanner
Rsector
Crystal
LSO BGO
ICATPP October 2009 U Pietrzyk - INM – FZ-Juelich - Germany Slide 31
GATE structure (2):
Geometry Constructionworld
Source
Body Head
Scanner
Rsector
Crystal
LSO BGO
ICATPP October 2009 U Pietrzyk - INM – FZ-Juelich - Germany Slide 32
GATE structure (2):
Geometry Constructionworld
Source
Body Head
Scanner
Rsector
Crystal
LSO BGO
ICATPP October 2009 U Pietrzyk - INM – FZ-Juelich - Germany Slide 33
GATE structure (2):
Geometry Constructionworld
Source
Body Head
Scanner
Rsector
Crystal
LSO BGO
ICATPP October 2009 U Pietrzyk - INM – FZ-Juelich - Germany Slide 34
GATE structure (2):
Geometry Constructionworld
Source
Body Head
Scanner
Rsector
Crystal
LSO BGO
ICATPP October 2009 U Pietrzyk - INM – FZ-Juelich - Germany Slide 35
GATE structure (3):
Two Complete Examples
Multi-ring PET
D. StrulIPHE Lausanne
Triple-head gamma camera
S. StaelensUni Ghent
D. StrulIPHE Lausanne
ICATPP October 2009 U Pietrzyk - INM – FZ-Juelich - Germany Slide 36
GATE structure (4):
Source Management
36
• Multiple sources controlled by source manager inserted via scripting complex geometries: customized GPS
(General Particle Source) Also: voxelized Sources, i.e. brain phantoms
ICATPP October 2009 U Pietrzyk - INM – FZ-Juelich - Germany Slide 37
GATE structure (5):
Timing and Motion
• Simulation time– a clock models the passing of
time during experiments– the user defines the
experiment timing• Time-dependant objects
– updated when time changes– allows programming of
movement, tracer kinetics...
ICATPP October 2009 U Pietrzyk - INM – FZ-Juelich - Germany Slide 38
GATE structure (6):
Physical Processes
PHOTONSELECTRONS
Standard photoelectric effect LE Compton scattering
Standard Gamma conversion
Standard Ionisation
Standard Bremsstrahlung
LE photoelectric effect
Standard Compton scattering LE Rayleigh scattering
LE Gamma conversion
LE Ionisation
LE Bremsstrahlung
• Choices of processes via scripting commands:low energy, standard or inactive
• Cut settings
ICATPP October 2009 U Pietrzyk - INM – FZ-Juelich - Germany Slide 39
GATE structure (7):
Sensitive Detectors / Digitizer
Hits
Digis
Energyresponse
Spatialresponse
Centroidreadout
ThresholdElectronic
s Pre-programmed components
– Sensitive detectors– Trajectory analyser
Digitizer– Linear signal processing
chain– Modular: set-up via scripting
ICATPP October 2009 U Pietrzyk - INM – FZ-Juelich - Germany Slide 40
GATE structure (8):
Data Output Formats• Multiple parallel output channels:
ROOT (real-time display, storage in ROOT files for further analysis)
ASCII files Binary files, incl. voxelized formats Specific scanner formats (e.g Crystal Clear LMF…)(Find ROOT at http://root.cern.ch/drupal/)
GATE simulation Sinogram Reconstructed image
ICATPP October 2009 U Pietrzyk - INM – FZ-Juelich - Germany Slide 41
Background of GATE:
41
• OpenGATE Collaboration founded in 2002• http://opengatecollaboration.healthgrid.org• Spokespersons:
• Christian Morel (Lausanne / Marseille; till 2003)
• Irene Buvat (INMC, Orsay / Paris; from 2003)GATE: a simulation toolkit for PET and SPECT
Phys. Med. Biol. 49 (2004) 4543–4561
S Jan1, G Santin2,24, D Strul2,25, S Staelens3, K Assie4, D Autret5, S Avner6, R Barbier7, M Bardies5, P M Bloomfield8, D Brasse6, V Breton9, P Bruyndonckx10, I Buvat4, A F Chatziioannou11, Y Choi12, Y H Chung12, C Comtat1, D Donnarieix9,13, L Ferrer5, S J Glick14, C J Groiselle14, D Guez15, P-F Honore15, S Kerhoas-Cavata15,
A S Kirov16, V Kohli11, M Koole3, M Krieguer10, D J van der Laan17, F Lamare18, G Largeron7, C Lartizien19, D Lazaro9, M C Maas17, L Maigne9, F Mayet20, F Melot20, C Merheb15, E Pennacchio7, J Perez21, U Pietrzyk21, F R Rannou11,22, M Rey2, D R Schaart17, C R Schmidtlein16, L Simon2,26, T Y Song12, J-M Vieira2, D Visvikis18, R Van deWalle3, EWieers10,23 and C Morel2
ICATPP October 2009 U Pietrzyk - INM – FZ-Juelich - Germany Slide 42
Background of GATE:
42
ICATPP October 2009 U Pietrzyk - INM – FZ-Juelich - Germany Slide 43
Outline:
• Intro: Neuroscience in the Research Center Jülich
• Motivation: Why using MonteCarlo Techniquesfor Simulation?
• Some Basics on Medical Imaging
• The Concept of GATE / GEANT4 / ROOT
• Examples of GATE Applications
• Summary & Demo
ICATPP October 2009 U Pietrzyk - INM – FZ-Juelich - Germany Slide 44
Clinical Example (I)
44
Estimating Scatter Contribution in Planar Gamma Camera Studieswith Iodine 131
A. Zakhnini (Diploma Thesis, University of Wuppertal)
GE Infinia 3/8’’ / HawkeyeHelios Clinic Wuppertal
ICATPP October 2009 U Pietrzyk - INM – FZ-Juelich - Germany Slide 45
Clinical Example (II)
45
Estimating Scatter Contribution in Planar Gamma Camera Studieswith Iodine 131
A. Zakhnini (Diploma Thesis,
University of Wuppertal)
Lateral View
ICATPP October 2009 U Pietrzyk - INM – FZ-Juelich - Germany Slide 46
Clinical Example (II)
46
Estimating Scatter Contribution in Planar Gamma Camera Studieswith Iodine 131
A. Zakhnini (Diploma Thesis,
University of Wuppertal)
Lateral View
ICATPP October 2009 U Pietrzyk - INM – FZ-Juelich - Germany Slide 47
Clinical Example (II)
47
Estimating Scatter Contribution in Planar Gamma Camera Studieswith Iodine 131
A. Zakhnini (Diploma Thesis,
University of Wuppertal)
Lateral View
ICATPP October 2009 U Pietrzyk - INM – FZ-Juelich - Germany Slide 48
Simulating new Developments for PET in GATE (I)
48
LSO
LSO
Classic solution:Scintillator + PMT
Modern solution:Scintillator + APD
More Compact!
PMT: Size:10-50 mmGain: up to 10**6Risetime: 1 nsQE: 20 %
APD:Size:5x5 mm**2Gain: up to 200Risetime: 5 nsQE: 70 %
ICATPP October 2009 U Pietrzyk - INM – FZ-Juelich - Germany Slide 49
Simulating new Developments for PET in GATE (II)
49
• more compact PET• much less “dead space”
• higher sensitivity
APDHamamatsu
4x8 elements10.5x20.7 mm2
pixelizedscintillator block
monolithicscintillator block
ICATPP October 2009 U Pietrzyk - INM – FZ-Juelich - Germany Slide 50
Simulating new Developments for PET in GATE (III)
5050
Simulation of Optical Photons in a
monolithic detector
N. Kobert, PhD-study, FZ-Juelich / Uni Wuppertal
APD(4 x 4
Elements)
Crystal (wrapped in Teflon) full of optical
photons
Incident Gamma (511 keV)
ICATPP October 2009 U Pietrzyk - INM – FZ-Juelich - Germany Slide 51
Simulating new Developments for PET in GATE (IV)
5151 N. Kobert, PhD-study, FZ-Juelich / Uni Wuppertal
ICATPP October 2009 U Pietrzyk - INM – FZ-Juelich - Germany Slide 52
Simulating new Developments for PET in GATE (V)
5252 N. Kobert, PhD-study, FZ-Juelich / Uni Wuppertal
ICATPP October 2009 U Pietrzyk - INM – FZ-Juelich - Germany Slide 53
Simulating new Developments for PET in GATE (VI)
5353
Simulation of Optical Photons in a monolithic
detector
N. Kobert, PhD-study, FZ-Juelich / Uni Wuppertal
Difference: Gamma coordinates (x / y)and centroid of optical photons
Centroids of Optical Photons (x / y)
Deviation x Deviation y
ICATPP October 2009 U Pietrzyk - INM – FZ-Juelich - Germany Slide 54
Simulating new Developments for PET in GATE (VII)
5454
Simulation of ClearPET Neuro
at FZ Juelichwith
GATE v 5
-----
Dynamic Simulationincorporating Rotation
of Gantry
ICATPP October 2009 U Pietrzyk - INM – FZ-Juelich - Germany Slide 55
Simulating new Developments for PET in GATE (VIII)
5555
ClearPET Neuro: RAT study
ICATPP October 2009 U Pietrzyk - INM – FZ-Juelich - Germany Slide 56
Simulating new Developments for PET in GATE (IX)
5656
Simulation of ClearPET Neuro:Source Distribution of a Homogeneous Cylinder
Rat Study with ClearPET NeuroReconstruction with STIR
ICATPP October 2009 U Pietrzyk - INM – FZ-Juelich - Germany Slide 57
Acknowledgements
57
Special thanks to Markus Axer (INM-1, FZ-Juelich)Alexandra Hellerbach (INM-1, FZ-Juelich)Natalia Kobert (INM-1, FZ-Juelich)Jürgen Scheins (INM-4, FZ-Juelich)Karl Ziemons (ZEL, FZ-Juelich)
Hamid Zakhnini (University of Wuppertal)Klaus Gasthaus (Helios Clinic Wuppertal)
&
all Members of the OpenGATE Collaboration
ICATPP October 2009 U Pietrzyk - INM – FZ-Juelich - Germany Slide 58
Contact
58
Prof. Dr. Uwe Pietrzyk(Physicist)Deputy DirectorInstitute of Neurosciences and Medicine (INM-1)Structural and Functional Organization of the BrainGroup Leader: Multi Modality Image Processing and MorphometryResearch Center Juelich GmbH, GermanyE-Mail: [email protected]://www.fz-juelich.de/INM&Department of Mathematics and Natural SciencesUniversity of Wuppertal, Germany