Nuclear Physics in MedicineChapter: Medical Imaging

NuPECC liaisons1Alexander Murphy and 2Faiçal Azaiez

1 The University of Edinburgh, UK2 IPN Orsay, IN2P3-CNRS, France

Conveners3Jose Manuel Udias and 4David Brasse

3 Universidad Complutense Madrid, Spain4 IPHC Strasbourg, IN2P3-CNRS, France

Piergiorgio Cerello, INFN Torino, ItalyChristophe de La Taille, Omega/IN2P3/CNRS, FranceAlberto Del Guerra, University of Pisa, ItalyNicola Belcari, University of Pisa, ItalyPeter Dendooven, University of Groningen, The NetherlandsWolfgang Enghardt, University Hospital TU Dresden, GermanyFine Fiedler, Helmholtz-Zentrum Dresden-Rossendorf, GermanyIan Lazarus, STFC, Daresbury Laboratory, Warrington, United KingdomGuillaume Montemont, CEA/LETI, FranceChristian Morel, CPPM/IN2P3/CNRS, Aix-Marseille University, FranceJosep F. Oliver, IFIC, Valencia University, SpainKatia Parodi, Ludwig Maximilians University Munich, GermanyMarlen Priegnitz, Helmholtz-Zentrum Dresden-Rossendorf, GermanyMagdalena Rafecas, IFIC, Valencia University, SpainChristoph Scheidenberger, Justus-Liebig-University Giessen and GSI-Darmstadt, GermanyPaola Solevi, IFIC,Valencia University, SpainPeter .G. Thirolf, Faculty of Physics at LMU Munich, GermanyIrene Torres-Espallardo, IFIC, Valencia University, Spain

List of Contributors

~ 10 years ago…

PET/CT is a technical evolution that has led to a medical revolution(Johannes Czernin, UCLA, 2003)

Invention of the Year

SNM Image of the Year

Anthony Stevens, Medical Options

From IMV

PET Clinical Procedures in US

…in Europe… (data from Anthony Stevens, Medical Options, EANM 2011)

In 2011,

Number of patient studies using PET or PET/CT: - between 2005 and 2010: 21 % increase - 2011: > 900 000 exams

FDG availability, scanner technology, … - 2010: 506 providers of PET or PET/CT in western Europe

64 %: public facilities

Average patient per scanner 2002 : 651 2010: 1559

PET Time Of Flight ImprovementFrom 400 ps to ….

From David Townsend (2008 AAPM Summer school)PET/MRI is a medical evolution based on a technical revolution (Thomas Beyer)

Nowadays…Few highlights

Spectral CT, K-edge imaging

Courtesy of C Morel et al, CPPM, France

Outline

• From Nuclear to Molecular Imaging– Small animal imaging system

• New Challenges– Detector design– Photon counting: towards spectral CT– -PET imaging– Simulation and reconstruction

• Interfaces– Quality Control in Hadrontherapy– Mass Spectrometry

From Nuclear to Molecular Imaging

The necessity of understanding biochemical processes at the molecular level

Advance in technological instrumentation

Preclinical Imaging

PET:the merging of biology and imaging

into molecular imaging(M Phelps)

University of Pittsburgh

Major efforts are devoted towards obtaining higherSensitivity

Spatial resolutionCheaper and easier to handle

15 MBq to target platelet,IPHC, Strasbourg

SPECT/CT PET/CT

SPECT/MR PET/MR

From Mediso Judenhofer et al, Nat. Med 14, 459-465, 2008

Courtesy of Dr. Piero A., Salvadori and Dr. Daniele Panetta, IFC-CNR Pisa

New Challenges…in Detector Design

• PET/CT Hybrid Imaging virtually available anywhere– Clinical routine in cancer staging, therapy assessment

• PET/MRI Hybrid Imaging… on its way

• Excellent performance

Can the performances be improved? Why?

• Better image quality and/or Lower dose• Better sensitivity & specificity in disease detection• Quantitative PET analysis

– that also requires protocol standardization

• Shorter Exam Time / Lower Cost

How to improve?

• 4D detectors with new design– Depth of Interaction, Time Of Flight– MR compatibility– Compactness– Cost & Scalability

How to Improve the Design ?

• Scintillators• Photon Detectors• Front-End Electronics• System Design & Integration

Dorenbos et col., IEEE TNS, 57, 2010 pp1162-1167

Y (ph/keV): 9Decay (ns): 300R (%): 10

Y (ph/keV): 30Decay (ns): 40R (%): 10

Y (ph/keV): 60Decay (ns): 16R (%): 3

How to Improve the Design ?

• Scintillators• Photon Detectors• Front-End Electronics• System Design & Integration

How to Improve the Design ?

• Scintillators• Photon Detectors• Front-End Electronics• System Design & Integration

• “Catch the first de-excitation photon”– Speed– Low Noise– Low (double) threshold– Low power consumption

Courtesy of Christophe De La Taille, Omega

How to Improve the Design ?

• Scintillators• Photon Detectors• Front-End Electronics• System Design & Integration

• Segmented / Continuous crystal• Radial/ axial orientation• Block structure / 1:1 coupling

System Performances - Spatial & timing resolutions - Count rate capability - Overall sensitivityCost/compactness/scalability

A lot of projects going on…Focus on the AX-PET collaboration

It consists only of two camera modules•48 LONG LYSO crystals (6 layers x 8 crystals)•156 plastic WLS strips (6 layers x 26 strips)

7

3.5

• Crystals are staggered by 2 mm. • Crystals and WLS strips are read out on alternate

sides to allow maximum packing density. • The other side is Al-coated, i.e. mirrored.

The layers are optically separated from each other.

Hamamatsu MPPC3×3 mm2

Hamamatsu MPPC3.22×1.19 mm2

3×3×100 mm3

3×0.9×40 mm3

Courtesy of the AXPET Collaboration

<RE>511 = 11.7 % (FWHM)1.48 mm FWHM in the axial direction

Photon Counting…towards spectral CT

Originally developed for vertex detectors in high energy physicsHybrid pixel arrays could replace conventional « charge integration »

Advantages: - absence of dark noise, - a high dynamic range - photon energy discrimination

-> Can provide spectral information

Pixelized sensorSi, CdTe, CZT

Readout ElectronicsStandard CMOS process

Technical specification of some hybrid pixel detector circuits

Photon Counting…towards spectral CT

Photon Counting…towards spectral CT

K-edge imaging of iodine

XPAD3 cameraXPAD-S ASIC500um thick silicon sensors500 kpixels130x130 um2 pixel pitch

Courtesy ofF Cassol Brunner and C Morel, CPPM, France

Medical Imaging using + Coincidences

whole class of potential PET isotopes excluded from medical application:

44mSc, 86Y, 94Tc, 94mTc, 152Tb, or 34mCl

3rd, higher-energy ray emitted from excited state in daughter nucleus:

- resulting extra dose delivered to the patient - expected increase of background from Compton scattering or pair creation

Perspective:

turn alleged disadvantage into promising benefit:

provided the availability of customized gamma cameras higher sensitivity for reconstruction of radioactivity distribution in PET examinations

PET imaging: so far (exclusive) + emitters: 18F, 11C

All present approaches towards ‘triple- imaging’ or ‘-PET’ :

based on Compton Camera:

Medical Imaging using + Coincidences

XEMIS: Xenon Medical Imaging System (since 2004 by Subatech, Univ. Nantes)

- cryogenic Time Projection Chamber (TPC) filled with liquid xenon (LXe) acting simultaneously as scatter, absorption and scintillation medium for the additional 3rd photon

E/E ~ 5.7% (511 keV 4.3% (1.157 MeV)

~ 1.25o

x = 2.3 mm (10 cm distance)

TPC

C. Grignon et al., Nucl. Instr. Meth. A 571 (2007) 142.T. Oger et al., Nucl. Instr. Meth. A 695 (2012) 125.J. Donnard et al., Nucl. Med. Rev. 15 (2012), C64–C67

Example: PET + TPC

New Challenges…in Simulation & Reconstruction

Both tomographic reconstruction and Monte-Carlo methods became feasible thanks to the advances in computer technology

The development of novel prototypes for emission tomography is usually supported by dedicated Monte-Carlo simulations and image reconstruction algorithms.

Monte-Carlo simulations are useful to optimize the system design and understand the observed phenomenaImage reconstruction is needed to determine the (expected) prototype performance at image level

Common Challenge

Model accuracy / Image qualityLow High

Simple modelsSimple simulated phantomsFew iterations of the recon

Complex modelsComplex simulated scenarii

Computational burdenShorter Longer

Efforts required to optimize balance between accuracy & computing time

Main Challenges…in Simulation (Emission Tomography)

To increase simulation speed without jeopardizing model accuracyParallel ImplementationImplementation in GPUsImplementation in FPGAs

To keeping pace with novel technologies and research scenariiFurther experiments and validation studies might be needed

Example of Model complexity

Which phenomena should be included?Light transport / Electron tracking / Voxelized phantomsTime-dependent phenomena:

Radioisotope decay / Phantom motionScanner rotation / Accidental coincidenceElectronic chain: pile-up, dead-time... Moving phantoms

Radiationtherapy + Imaging Scenarios

Main Challenges…in Reconstruction

From Scanner to Image

+ =

Instrumentation Image Reconstruction

PhysicsTowards improving

image quality

Main Challenges…in Reconstruction

Originally:A 2D image representing radioisotope distribution within one section of the body

Nowadays:Reconstruction of 3D images (volume)Dynamic reconstruction (4D): time sequences

Recent advances: 5D and 6D reconstructionTime evolutionHeart/respiratory motionsKinetic parameters

Some examples…

D. Wiant et al. Med. Phys. 37. 2010

Modeling of the PSF

Clinical PET

Analytic vs iterative

Small animal PET

Interfaces…Quality Control in HadrontherapyMotivation: range uncertainties

Source: HZDR, DKFZ

A monitoring of the dose delivery is requiredIn order to fully profit from the advantages of ion beams

The range of the particlesthen the maximum dose deliveryis very sensitive to modifications:

tissue density,inaccuracies in patient positioning

Deviations in dose distribution

Interfaces…Quality Control in Hadrontherapy« Several methods of medical imaging in particle ion beam therapy are under investigation in order to measure the range of the particles in the tissue or even directly measure the applied dose in vivo »

Positron Emission Tomography

Prompt gamma ray imaging

Charged particles imaging

Ion radiography and tomography

Three implementations are investigatedIn-beam PET (GSI, NIRS, Catana)In-room PET (MGH, Kashiwa)Offline PET (HIT, Hyogo)

Different detector conceptsCollimated gamma cameraMulti slit cameraCompton cameraPrompt gamma timing

Recent proof-of-principle simulation and experimental studies reported from research group in France, Italy and Germany

Direct measurement of the residual range of high-energy low-intensity ions traversing the patient. Prototypes are under development for both protons and carbon ion beams

Interfaces…Quality Control in HadrontherapyFocus on PET: « the only clinically investigated method »

PET activation (right) measured after delivery of the planned carbon ion treatment dose (left) at HIT, in comparison to the corresponding PET MC prediction (middle). The arrow marks an example of good range agreement (adapted from [Bauer 2013] with permission).

On going developments: - TOF PET with T<200ps - Characterization of nuclear reaction cross sections - Feasibility of PET verification for moving targets - Extension to others ions - Solution for automated PET range evaluation in clinical routine - Application of high energy photon therapy

Interfaces…Mass Spectrometry

« Imaging Mass Spectrometry, where high spatial resolution is combined with mass spectrometric analysis of the sample material, is a versatile and almost universal method to analyze the spatial distribution of analytes in tissue sections”

Some examples:Tissue recognitionDrug developmentMultimodal imaging

Left: single-pixel mass spectrum of the outer stripe outer medulla; The green label indicates the mass peak that is characteristic for imatinib. Right: imaging mass spectrometry yields the distribution of different substances in the mouse kidney (figures reprinted from ref.  [Röm13]).

IMS spectra of a mouse kidney after treatment with the anti-cancer drug imatinib.

OutlookMedical imaging in general, and nuclear medicine in particular, has experienced and continues to exhibit evolution at exponential speeds.

The work performed in nuclear physics groups such as radiation detection, simulations, electronics, and data processing, find application in nuclear medicine.

This chapter provided a glimpse of how nuclear physics research has been involved in the advance of medical imaging and, more interestingly, how our current efforts are paving the way for the imaging technologies of tomorrow.

This chapter reflects the fact that inside the nuclear physics community, research and development activities in medical imaging detector development coexist, at times even within the same research group.

It is our duty to help and promote the translation of developments from our nuclear physics laboratories and basic nuclear science experiments into practical tools for the clinical and preclinical environments.

Piergiorgio Cerello, INFN Torino, ItalyChristophe de La Taille, Omega/IN2P3/CNRS, FranceAlberto Del Guerra, University of Pisa, ItalyNicola Belcari, University of Pisa, ItalyPeter Dendooven, University of Groningen, The NetherlandsWolfgang Enghardt, University Hospital TU Dresden, GermanyFine Fiedler, Helmholtz-Zentrum Dresden-Rossendorf, GermanyIan Lazarus, STFC, Daresbury Laboratory, Warrington, United KingdomGuillaume Montemont, CEA/LETI, FranceChristian Morel, CPPM/IN2P3/CNRS, Aix-Marseille University, FranceJosep F. Oliver, IFIC, Valencia University, SpainKatia Parodi, Ludwig Maximilians University Munich, GermanyMarlen Priegnitz, Helmholtz-Zentrum Dresden-Rossendorf, GermanyMagdalena Rafecas, IFIC, Valencia University, SpainChristoph Scheidenberger, Justus-Liebig-University Giessen and GSI-Darmstadt, GermanyPaola Solevi, IFIC,Valencia University, SpainPeter .G. Thirolf, Faculty of Physics at LMU Munich, GermanyIrene Torres-Espallardo, IFIC, Valencia University, Spain

List of Contributors