The Role of Silicon Radiation Sensors and Integrated Front-End Electronics In Medical Imaging Instrumentation.

Similar documents
CHIPP Plenary Meeting University of Geneva, June 12, 2008 W. Lustermann on behalf of the AX PET Collaboration

Development of a High Precision Axial 3-D PET for Brain Imaging

Compton Camera. Compton Camera

Compton Camera with PositionSensitive Silicon Detectors

Time-of-Flight PET using Cherenkov Photons Produced in PbF 2

Detector technology. Aim of this talk. Principle of a radiation detector. Interactions of gamma photons (gas) Gas-filled detectors: examples

Development of High-Z Semiconductor Detectors and Their Applications to X-ray/gamma-ray Astronomy

Radionuclide Imaging MII Positron Emission Tomography (PET)

Dual Isotope Imaging with LaBr3:Ce Crystal and H8500 PSPMT

MEDICAL EQUIPMENT: NUCLEAR MEDICINE. Prof. Yasser Mostafa Kadah

Development of HPDs. for applications. in physics and medical imaging

PoS(TIPP2014)033. Upgrade of MEG Liquid Xenon Calorimeter. Ryu SAWADA. ICEPP, the University of Tokyo

A dual scintillator - dual silicon photodiode detector module for intraoperative gamma\beta probe and portable anti-compton spectrometer

Radionuclide Imaging MII Detection of Nuclear Emission

Application of Nuclear Physics

PET. Technical aspects

Towards Proton Computed Tomography

Mayneord-Phillips Summer School St Edmund Hall, University of Oxford July Proton decays to n, e +, ν

A. I, II, and III B. I C. I and II D. II and III E. I and III

A Brief Introduction to Medical Imaging. Outline

ISPA-Tubes with YAP:Ce Active Windows for X and Gamma Ray Imaging.

Using new digital SiPM from Philips with AX-PET a new geometrical concept for PET

The Compton Effect. Martha Buckley MIT Department of Physics, Cambridge, MA (Dated: November 26, 2002)

DEVIL PHYSICS THE BADDEST CLASS ON CAMPUS IB PHYSICS

Prospects for achieving < 100 ps FWHM coincidence resolving time in time-of-flight PET

Development of a new MeV gamma-ray camera

Recent advances and future perspectives of gamma imagers for scintimammography

hν' Φ e - Gamma spectroscopy - Prelab questions 1. What characteristics distinguish x-rays from gamma rays? Is either more intrinsically dangerous?

Radiation Detection and Measurement

Detector R&D in KAPAC

The Hermes Recoil Silicon Detector

Tests of the BURLE 64-anode MCP PMT as the detector of Cherenkov photons

EEE4106Z Radiation Interactions & Detection

Initial Studies in Proton Computed Tomography

Chapter 4 Scintillation Detectors

Radioisotopes in action. Diagnostic application of radioisotopes. Steps of diagnostic procedure. Information from various medical imaging techniques

First Results and Realization Status of a Proton Computed Radiography Device

Lecture 16 Light transmission and optical detectors

Lecture 5: Tomographic nuclear systems: SPECT

III. Energy Deposition in the Detector and Spectrum Formation

Scintillation Detector

Position Sensitive Germanium Detectors for the Advanced Compton Telescope

arxiv: v1 [physics.ins-det] 29 Jun 2011

Measurement of the transverse diffusion coefficient of charge in liquid xenon

Positron Emission Tomography (PET)

Performance of the Gamma-ray Imaging Detector with Micro-TPC

Positron Emission Tomography

6: Positron Emission Tomography

Sample Spectroscopy System Hardware

Factors Affecting Detector Performance Goals and Alternative Photo-detectors

A RICH Photon Detector Module with G-APDs

FXA UNIT G485 Module X-Rays. Candidates should be able to : I = I 0 e -μx

Module of Silicon Photomultipliers as a single photon detector of Cherenkov photons

Final report on DOE project number DE-FG07-99ID High Pressure Xenon Gamma-Ray Spectrometers for Field Use

Light ion recoil detector

Development of semiconductor imaging detectors for a Si/CdTe Compton camera

Radiation Detector 2016/17 (SPA6309)

Position sensitive detection of thermal neutrons with solid state detectors (Gd Si planar detectors)

Measurements of liquid xenon s response to low-energy particle interactions

Positron-Electron Annihilation

Technical University of Denmark

Timing and Energy Response of Six Prototype Scintillators

1st Faculty of Medicine, Charles University in Prague Center for Advanced Preclinical Imaging (CAPI)

Drift plane. substrate (20ÉIm polyimide) 200ÉIm. Back strip (180ÉIm width) Base (Ceramic) Anode strip (10ÉIm width) Cathode strip (100ÉIm width)

Detector R&D at KIPAC. Hiro Tajima Kavli InStitute of Particle Astrophysics and Cosmology

A POSITION SENSITIVE ALPHA PARTICLE DETECTOR BASED ON A LYSO CRYSTAL AND A MICRO-PIXEL AVALANCHE PHOTODIODE

Queen s University PHYS 352

List of Nuclear Medicine Radionuclides. Nuclear Medicine Imaging Systems: The Scintillation Camera. Crystal and light guide

SCINTILLATION DETECTORS AND PM TUBES

DEPFET sensors development for the Pixel Detector of BELLE II

Nuclear Instruments and Methods in Physics Research A

arxiv: v2 [physics.ins-det] 8 Feb 2013

Detector R&D at KIPAC

Chicago PET Development and

There are three mechanisms by which gamma rays interact with absorber atoms from which two are important for nuclear medicine.

Nuclear Medicine Intro & Physics from Medical Imaging Signals and Systems, Chapter 7, by Prince and Links

Scintillation Detectors

Introduction to SPECT & PET TBMI02 - Medical Image Analysis 2017

X- & γ-ray Instrumentation

PHOTODETECTORS AND SILICON PHOTO MULTIPLIER

Neutron Induced Nuclear Counter Effect in Hamamatsu Silicon APDs and PIN Diodes

Gamma-ray spectroscopy with the scintillator/photomultiplierand with the high purity Ge detector: Compton scattering, photoeffect, and pair production

Radioactivity. Lecture 6 Detectors and Instrumentation

Performance of high pressure Xe/TMA in GEMs for neutron and X-ray detection

Quantitative Assessment of Scattering Contributions in MeV-Industrial X-ray Computed Tomography

A NEW GENERATION OF GAMMA-RAY TELESCOPE

A gas-filled calorimeter for high intensity beam environments

Readout of LYSO using a new silicon photodetector for positron emission tomography

Gamma-ray spectroscopy with the scintillator/photomultiplierand with the high purity Ge detector: Compton scattering, photoeffect, and pair production

Structure of Biological Materials

The Silicon-Tungsten Tracker of the DAMPE Mission

A novel design of the MeV gamma-ray imaging detector with Micro-TPC

Calorimeter test-beam results with APDs

Development of a 3D-Imaging Calorimeter in LaBr 3 for Gamma-Ray Space Astronomy

Tests of the BURLE 64-anode MCP PMT as the detector of Cherenkov photons

Charge collection in PET detectors

Detection and measurement of gamma-radiation by gammaspectroscopy

Technical University of Denmark

1 Introduction. KOPIO charged-particle vetos. K - RARE Meeting (Frascati) May Purpose of CPV: veto Kl

Photon Instrumentation. First Mexican Particle Accelerator School Guanajuato Oct 6, 2011

Transcription:

The Role of Silicon Radiation Sensors and Integrated Front-End Electronics In Medical Imaging Instrumentation. SLAC Seminar 2. May 2007 P. Weilhammer INFN Perugia/CERN 3. 05. 2007 Seminar SLAC 1

OUTLINE of Talk 1. Overview of present Medical Imaging Modalities 2. Photon Detection with Silicon Radiation Detectors and Implications for Applications of Silicon in Medicine 3. Examples of Silicon Detectors in Medical Imaging Applications 3. 05. 2007 Seminar SLAC 2

The Famous Picture 3. 05. 2007 Seminar SLAC 3

Medical Imaging in the 21 st Century: Multi Modality 3. 05. 2007 Seminar SLAC 4

While photographic emulsion was for a very long time without competition, important innovations in medical imaging were introduced over the last 50 to 60 years: Anger Camera for SPECT First attempts on PET with Proportional Wire Chambers PET Scanners with High Z Scintillators and PM Readout using Anger Logic MRI Attempts on Electronic Collimation in SPECT using Compton Scattering of the Gamma Ray with Germanium Detectors XR-CT with Silicon Photo Diode arrays in Current Mode Ultrasound Scanners Etc.. 3. 05. 2007 Seminar SLAC 5

With the exception of X-Ray CT where Silicon Photo Diodes Play a dominant role, Silicon Radiation Sensors are not very strongly represented in this field so far. The dominant detector technologies are Scintillators PM Tubes Maybe soon High Z semiconductors in Digital X-Ray CT One exception: Low Dose digital Mammography Scanner from SECTRA (single sided strip detectors edge on and VLSI Front-end) See www.sectra.com Very promising for Screening in Mammography 3. 05. 2007 Seminar SLAC 6

Medical Imaging Modalities A short incomplete list of imaging modalities which might be improved by the implementation of silicon detectors: 1. Field of X-ray imaging X-ray radiography; projection images to obtain 2 D anatomical information. Classical X-ray images X-ray Computed Tomography; 3-D anatomical information through reconstruction of the distribution of attenuation coefficient μ(x,y,z); Tissue specific contrast is obtained by measuring at the detector I = I 0 exp(-μ(x,y,z) d) Most often used in patient diagnostics in all hospitals: Translation-Rotation Scanner. Measure many slices in one or several rotations. Modern scanners have up to 64 slices. Radiation detectors are pixelated matrices of suitable (high Z) scintillators like CsI, BGO, LSO, more recently ceramic based scintillators. 3. 05. 2007 Seminar SLAC 7

Preclinical X-ray CT: small animal scanners. Similar to Clinical scanners with emphasis on higher spatial resolution for lower density tissue. High intensity X-ray sources are used with micro-focus or very high resolution synchrotron source. X-ray energy ranges: ~ 10 kev to 120 kev 2. Nuclear Medicine Imaging Gamma Camera Single Photon Emission Computed Tomography (SPECT) Positron Emission Computed Tomography (PET) High energy γ rays penetrate tissue with little absorption. Imaging is performed by injection and take-up in the patient of a radio-ligand containing a meta-stable reporter radionuclide which emits γ-rays or positrons and often also electrons which are absorbed in the surrounding tissue. He gamma rays will exit from the body with occasional Compton scatter. 3. 05. 2007 Seminar SLAC 8

The goal is to reconstruct the distribution of radioactivity within the body, either 2-D or 3-D image, using back projection algorithms. Traditional detectors are highly segmented scintillation crystal arrays (NaI, CsI, BGO, LSO, La-bromides, ). The scintillation arrays are readout by Photo Multipliers. Anger Camera: The direction of the photon absorbed in the detector is determined by the x,y-coordinate of the impact and by tight collimation in front of the scintillator (Pb-collimator with many small holes). Pin-Hole camera: the collimator is replaced by a arrangement of one or several specially shaped pinholes in a collimator structure. SPECT Camera PET Scanner; both SPECT and PET cameras allow direct recording of 2-D projections simultaneously (or by rotating one camera head around the patient in case of SPECT) leading to full 3-D image reconstruction (not slice by slice). Closed ring detector (scintillator) detector geometry to measure 180 degree 511 kev photons from positronium annihilation. Data Recorded are sinograms 3. 05. 2007 Seminar SLAC 9

Preclinical PET : high resolution small animal PET. Compton camera and Compton probes Autoradiography Bio-molecular Imaging is emerging.. In all imaging detectors and systems the important quality factors are: Detective Quantum Efficiency (DQE) (Sensitivity) Spatial resolution Speed, coincidence window 3. 05. 2007 Seminar SLAC 10

Detection of Photons and Energetic Electrons in Semiconductor Detectors Medical imaging requires good ability of detection of photons, in reality detection of energetic electrons created inside the material ( an advantage!), over a wide range of energies. Energy Ranges: Computed Tomography (CT) X-rays: 20 to ~>120 kev Single Photon Emission Tomography (SPECT): detect γ-rays for a big variety of isotopes used in different tracer molecules 99m Tc 111 In 31 I 140 kev 185 and 245 kev 360 kev Positron Emission Tomography: 511 kev γ from e + e - annihilation Autoradiography: β particles emitted from e.g. Tritium, 14 C, 33 Ph, ( from 10 kev to several 100 kev) 3. 05. 2007 Seminar SLAC 11

Photon Interactions in Silicon Only two out of all photon interactions are important for medical imaging: The wanted one: Photoelectric Absorption (total absorption of γ or X-ray) σ = 4 2 α 4 Εγ 7/2 Ζ 5 σ Th with the Thomson cross-section σ Th. = 8π/3 r 2 0 = 6.652 bars per electron. 3. 05. 2007 Seminar SLAC 12

3. 05. 2007 Seminar SLAC 13 The unwanted one: Compton scattering ) cos (1 1 2 θ γ γ γ + = c m E E E e + = ) cos (1 1 1 1 2 θ γ γ m c E E E e e θ The recoil electron ( from K-shell or L-shell or valence band) creates (eh) pairs in the semiconductor bulk through ionization Kinetic energy of recoil electron

Attenuation of incoming photons in material Good photon detector are detectors which absorb most of the incoming photons preferably by photoelectric absorption. Quantitatively the attenuation in the material of a sensor is characterized by the mass attenuation coefficient μ(e) : N No [ μ( E) / ρ] ( ρ t) μ( E) t = = e e With t the thickness of the sensor in direction of the photon beam and ρ the density of the material. Materials with low Z (silicon has Z=14) become quickly impractical with increasing photon energy! 3. 05. 2007 Seminar SLAC 14

In 1mm thick silicon for 20 kev photons Photoelectric interaction: ~ 97% Compton interactions: ~3% Interesting region for medical imaging Interactions/m for Si versus photon energy 3. 05. 2007 Seminar SLAC 15

Range of Electrons in Materials The range of electrons in materials expressed as range * density is very similar for many different materials Typical Range: 50 kev electron in silicon: ~20 μm NaI 200 kev : ~200 μm Range*density [g/cm 2 ] 500 kev :~ 600 μm 10-2 Si For Compton interaction the point-like domain is between 10 kev and 250 to 300 kev! 100 kev 3. 05. 2007 Seminar SLAC 16

Some inherent physical limitations in different imaging modalities are: Spatial extension of the photon interaction in the detector material due to the nature of photon interactions (in most materials interaction cascades are frequent before final absorption). The typical extension of a photon interaction in many detector materials ( at 500 kev) can be considered to be confined in a sphere of ~1 cm in diameter. Uncertainty in Depth of Interaction Parallax error Finite path length of positrons and recoil electrons Compton scattering in tissue. In PET: Finite momentum of e + e - compound at the moment of decay Acolinearity Accidental coincidences. ~1cm through multiple interactions in scintillator Incoming γ 3. 05. 2007 Seminar SLAC 17

Types of silicon radiation sensors which could be interesting in medial applications: Si strip detectors, single sided and double sided. Si pixel detectors Si pad detectors and micro pad detectors CMOS imagers Flat panel devices based on amorphous silicon Sub-micron fast CMOS front-end chips for readout of strips, pixels and pads and others Silicon Photo Multipliers(SiPM) Variety of Front-end deep submicron circuits developed for HEP 3. 05. 2007 Seminar SLAC 18

Advantages of Silicon Detectors over classical Instrumentation in Medical Imaging? For Si pixel, pad and strip detectors Very high segmentation feasible Matching of segmentation of front-end readout electronics Excellent energy resolution Excellent position resolution Possibility of using counting mode with energy weighting Low voltage operation.. 3. 05. 2007 Seminar SLAC 19

In the following I want to discuss possible applications of Silicon Detectors and Front-End Electronics ASICS which are projects within the CIMA Collaboration. Emphasis will be on New Developments for PET Compton Camera SPECT and Compton PET Micro X-Ray CT 3. 05. 2007 Seminar SLAC 20

R&D Projects using Silicon Detectors in Medical Imaging within the CIMA Collaboration Novel axial brain PET Scanner using Hybrid Photon Detectors (HPD) Readout of z-coordinate of Axial PET with Wave Length Shifters and Silicon Photo-Multipliers Compton Imaging and Probes High resolution small animal PET scanner based on Compton interactions A High Resolution Micro-CT Prototype Module for Small Animal Imaging 3. 05. 2007 Seminar SLAC 21

The CIMA Collaboration: Institutes Lisbon INFN Bari INFN Rome INFN Perugia HUG Geneva Phys. Dept. Uni Geneva University of Michigan University of Ljubljana Ohio State University LANL University of Valencia Karlsruhe Kharkhov Space Institute CERN Cracow Industrial Partners: IRST Trento Gamma-Medica_IDEAS SINTEF 3. 05. 2007 Seminar SLAC 22

New Development for PET 3. 05. 2007 Seminar SLAC 23

Some of the Shortcomings of Present Day Clinical PET Scanners A reference for the new generation of PET scanners could be the High Resolution Research Tomograph (HRRT) developed by CPS innovations * Efficiency for the detection of photon pairs is given to be 6.9%, which includes a sizeable fraction of unidentified Compton interactions. Energy resolution is 17% at 511 kev, timing resolution is 2 to 4 ns. The volumetric voxel resolution is given as 20 mm3, corresponding to a transaxial resolution of in average 2.6 mm and an axial resolution of about 3mm. All quantities referenced here are FWHM. * K. Wienhard et al, IEEE Trans. Nucl. Sci. 49 (2002) 104-110 3. 05. 2007 Seminar SLAC 24

The main shortcomings are: Relatively low efficiency of photon conversion due to the anticorrelation between accurate knowledge of depth of interaction and thickness of scintillation crystals. Parallax error due to limited knowledge of the depth of interaction in the radial direction of the 511 kev photon. Several techniques have been developed to reduce the parallax error, e.g. the Phoswich arrangement of scintillation crystals [HRRT]. Image smearing due to the physics of the photon interaction. The spatial extension of the 511 kev photon interaction cascade even in high Z scintillation material gives an important contribution to deteriorate the image quality. This is due to the fact that even for scintillation material with the highest density and effective Z the fraction of Compton interactions is 60% or more. Limited capability to identify and reject events with a Compton interaction in the scintillation material. 3. 05. 2007 Seminar SLAC 25

Some ideas to improve over present day scanners. 3. 05. 2007 Seminar SLAC 26

Novel Axial Geometry PET Scanner A proposal for a parallax-free Compton enhanced PET camera module for high resolution, high sensitivity functional brain imaging based on a Hybrid Photon Detector (PET-HPD) 3. 05. 2007 Seminar SLAC 27

The HPD hν Bi-alkali photocathode Perfect single photon detection 1 ph.e. 2 ph e e - 12 σ 3 ph e Si Sensor 2048 pads (1 x 1 mm 2 ) 32 28 24 HPD PC87 (produced Easter Sunday 2001) 16 front-end chips Ceramic PCB Q.E. (%) 20 16 12 8 4 0 200 300 400 500 600 lambda (nm) 3. 05. 2007 Seminar SLAC 28

This concept is discussed in detail in J. Seguinot et al., Il Nuovo Cimento C, Vol. 29 Issue 04 pp 429-463 3. 05. 2007 Seminar SLAC 29

The Concept Principle of a camera module y z x HPD1 HPD2 3. 05. 2007 Seminar SLAC 30

Discriminate Compton interactions: Fine 3D segmentation makes it possible unambiguous ambiguous γ reconstruction point which was the point of 1 st γ interaction? γ γ Select only events in which Compton scattering happens in forward hemisphere Restrict to Compton angle 10 θ 60 Ask for energy deposit in first interaction E 170 kev Throw out of data sample Compton events which cannot be resoved w.r.t. first vertex 3. 05. 2007 Seminar SLAC 31

x and y resolution :axially arranged, long LYSO scintillation crystal bars allow to choose x and y resolution according to the chosen lateral dimension (s) of the bars. z-resolution : optimize light yield N1 and N2, read on both sides by HPDs and light absorption along the bars. A well optimized bulk absorption in the scintillation bars and highest possible light yield are the most important parameters. 3. 05. 2007 Seminar SLAC 32

The z coordinate and the spatial resolution in z is Monte Carlo simulations showed that a good compromise for LYSO crystals is a crystal length of 15 cm with λ around 100 mm. These simulations indicate that σ z = 3.5 4.5 mm could be obtained The axial resolution is not as good as one would desire for an ideal instrument 3. 05. 2007 Seminar SLAC 33

The Hybrid Photon Detector: PET-HPD Some relevant properties of HPD for PET application: Very good spatial resolution can be chosen; size, geometry and granularity of silicon pad sensor can be chosen according to the requirements. Very good energy resolution charge gain in a single stage dissipation Very good linearity over large dynamic range Not temperature sensitive 11. May 2005 EUROMEDIM2006 Marseille 34

A Prototype PET HPD has been successfully built and tested 3. 05. 2007 Seminar SLAC 35

Basic Elements: Double metal pad detectors and VATAGP5 Chip 3. 05. 2007 Seminar SLAC 36

Some sensor properties: Full depletion voltage: V ~ 30 V Leakage current per pad: ~500 pa Pad and routing line cap.: ~ 5 pf Self-triggering Front-End Electronics: the VATAGP5 chip Fast Charge Sensitive Preamplifier Output of preamp fanned out to Slow shaper amplifier (t=220nsec) followed by a S/H to record precisely pulse height (energy) Fast shaper (t=40 ns) followed by a discriminator with time walk compensation and a monostable; firing of discriminator initiates a S/H Repeat pattern 128 times; all 128 channels have common threshold 3. 05. 2007 Seminar SLAC 37

VATAGP5 continued 3-bit trim DAC for trimming thresholds for each channel individually The mono-stable pulse initiates readout clock and S/H Four readout modes Serial reads sequentially S/H of all channels Sparse puts address of hit channel(s) into a register and only those channels will be read Sparse with neighbors reads hit channel and n neighbors. Sparse with any pre-defined neighbors Dynamic range: up to 1.2 pc for positive polarity. 3. 05. 2007 Seminar SLAC 38

Results from tests with first Prototype PET-HPD 3. 05. 2007 Seminar SLAC 39

First Brain PET HPD works: Hit Distribution from Light Spot This lego plot shows that a threshold of ~ 20 fc eliminates easily any dark current hits Background free images 3. 05. 2007 Seminar SLAC 40

μ (ADC counts) 500 400 300 200 100 Mean Charge μ and σ/μ of charge distributions as Function of Cathode Voltage σ/μ chan. #45 chan. #53 0 4.0 5 7 9 11 13 15 17 19 21 V PC (kv) Mean charge μ (left axis) and ratio of Gaussian width to mean charge σ/μ (right axis) versus cathode voltage UC (kv). μ 3. 05. 2007 Seminar SLAC 41 7.0 6.5 6.0 5.5 5.0 4.5 σ/μ (%) Note: that mean charge is linear but intercepts at ~6 kev due to energy loss in dead layer. This can be improved in next sensor production run. Expect intercept at 0.5 kev, which will considerably improve the charge gain in the HPD σ/μ reaches almost a plateau around 17 kev since energy straggling becomes small. One can estimate an energy loss of ~ 1.6 kev at 20 kv with nearly negligible straggling. Gain at 20 kv is 5090 From σ/μ = (ENF/N) 1/2 N = 507 photo electrons( ~ N 0 of LYSO) The absolute gain of the chain can now be calculated: 0.94 fc/(adc count) for chip1

Timing is another crucial problem: Time walk as a function of distance from threshold for Cr-RC shaping: FWHM time resolution versus Threshold Distance One can obtain ~ 5 nsec FWHM with VATAGP-5 ASIC Number of times threshold for coincidence window 3. 05. 2007 Seminar SLAC 42

Status A first PET HPD tube for readout of a scintillation crystal matrix developed for use in a novel axial PET concept has been designed, fabricated and tested. All the relevant features of the mechanical and optical properties of the envelope, the front-end electronics chip and the silicon pad sensors, required for this application have been successfully demonstrated. The system has an appropriate dynamic range which will allow detecting energy deposits from 30 kev to well above 511 kev in a LYSO crystal with very good linearity. The required time resolution (~ 5ns FWHM) needed for PET can be achieved with the VATAGP-5 ASIC. The fabrication of a second PET-HPD tube is under way. The next major step in the project is to assemble a complete camera module to characterize its spatial (axial coordinate) and energy resolution 3. 05. 2007 Seminar SLAC 43

A New Concept To Obtain Optimal Axial Spatial Resolution 3. 05. 2007 Seminar SLAC 44

The Principle: LYSO Crystal Bar Thin WLS Strip Question: is there enough light from WLS strips for 511 kev photon in LYSO or even for 50 to 100 kev Compton recoil electrons? 3. 05. 2007 Seminar SLAC 45

For a Brain Pet: Crystal size: 3 mm x 3 mm x 150 mm WLS size: 1 mm x 3 mm x 35 mm Measurement of z-coordinate: either take WLS strip with highest hit σ z = 3mm/Sqrt(12) = 0.9 mm Or measure analog values on more than one strip: center of gravity; should in general be better than digital resolution 3. 05. 2007 Seminar SLAC 46

Experimental Verification: Two different and independent methods to establish validity of new concept Adjustable pulsed low energy electron beam 22 Na source in coincidence 3. 05. 2007 Seminar SLAC 47

Measure Photoelectric yield of WLS strips Achievable spatial resolution along the crystal axis Timing resolution 3. 05. 2007 Seminar SLAC 48

Measured charge distribution in each of the 2 WLS strips with the electron beam moved across the WLS strips (~350 kev) 3. 05. 2007 Seminar SLAC 49

With the existing set-up one can generate a charge in the LYSO equivalent to a photon energy of up to 400 kev. Extrapolate curve below to 511 kev measured photo electric yield in WLS is 42 p.e. The WLS were read with Hamamatsu PM s with 15% Photo efficiency. In a real device one will use SiPM for WLS readout; Hamamtsu quotes 40% yield for their MPPC; this will give ~ 100 p.e. in two WLS strips Signals from both strips summed 3. 05. 2007 Seminar SLAC 50

Experimental results for z spatial resolution In this setup always 2 WLS strips hit: calculate z with formula: Correlation between beam spot and measured z- coordinate Beam moved across the 2 WLS strips 3. 05. 2007 Seminar SLAC 51

Z-resolution is dominated by the p.e. statistics in the WLS strips. Expect a 1/SQRT(E conv ) dependence. Z-resolution as a function of electron beam energy The real value of sigma(z) after deconvolution of beam spot size is ~800micron 3. 05. 2007 Seminar SLAC 52

GM-APD Arrays (just a few words before discussing results with SiPM Set-Up) Commercial SiPM from Hamamatsu S. Uozumi, Talk at VCI, 2007 3. 05. 2007 Seminar SLAC 53

A full wafer with Si-PM structures ; produced by IRST, Trento,Italy Waf er Main block 3. 05. 2007 Seminar SLAC 54

Our Measurements IRST 1mm x 1mm SiPM read with P/N MSA 0886- BLK HP fast Ampl., Single Photon Response V bkd = 35 V, Vbias = 38 V Gain = 1.25 x 10 6 Pedestal 8000 7000 Plot 032 1 photon 6000 No. of counts 5000 4000 3000 2000 1000 0-5.00E-009-4.00E-009-3.00E-009-2.00E-009-1.00E-009 0.00E+000 1.00E-009 Amplitude 3. 05. 2007 Seminar SLAC 55

Single Photon Response of Hamamatsu 3mm x 3mm SiPM Rad with Fast HP Ampl. V bkd = 69.7 V, V bias = 71.2 V, Gain 6.7 x 10 5 350 Plot 031 300 250 No.of counts 200 150 100 50 0-3.00E-009-2.00E-009-1.00E-009 0.00E+000 1.00E-009 Amplitude 3. 05. 2007 Seminar SLAC 56

Results from the second method using 22 Na source and G- APD (Hamamtsu MPPC) readout on LYSO and on WLS strip Pulse Height Spectrum with G-APD and LYSO: ΔE/E ~12% 3. 05. 2007 Seminar SLAC 57

Measured Pulse Height Spectrum fro WLS with G-APD (For photo-electric events) ~ 35-40 p.e. 3. 05. 2007 Seminar SLAC 58

Timing of LYSO w.r.t. WLS Strip with G-APD: FWHM ~700ps 3. 05. 2007 Seminar SLAC 59

The concept of axial (z) coordinate measurements using a WLS strip matrix looks very promising for PET imaging: Next step is the construction of two prototype crystal stacks with WLS matrix readout. Test both HPD and SiPM readout of LYSO 3. 05. 2007 Seminar SLAC 60

This novel concept could solve most of the problems inherent in present day PET systems. Summary of (expected) Performance: Full 3D reconstruction of the 511 kev photons No parallax error Spatial resolution (x, y, and z) can be chosen according to requirements by selecting Crystal and WLS dimensions. The total thickness of the scintillation detector stack can be chosen independently of other device parameters, which allows in principle to choose the efficiency according to the requirements of specific applications. Total uniformity of spatial resolution over the complete field of view. Capability to distinguish photon interactions with Compton cascades from photoabsorption events with nearly 100% efficiency.. 3. 05. 2007 Seminar SLAC 61

Increase of sensitivity by including in the final event sample events with a primary Compton interaction exploiting the constraints given by energy deposited in the scintillation crystals and the position measurement of both observed interactions. About 25% of the Compton interactions can be kinematically fully resolved. This will increase for LYSO the number of coincidences to be used for chord reconstruction by a factor 1.6 to 1.8, depending on the recoil electron energy cut-off. Axial arrangement of the scintillation crystals can reduce the number of electronic readout channels, while maintaining high granularity. Very good energy resolution in the order of 8% at 511 kev if the LYSO crystal matrix is readout with a HPD and ~12% with SiPM readout. Competitive timing resolution of ~700ps for SiPM readout of LYSO, maybe also useful for TOF PET. The spatial resolution which can be obtained with the new concept (for scintillation crystal dimensions proposed in[5]) will result in a voxel precision of 9 mm 3 FWHM, close to the limitations imposed by the inherent physical limits from a-co-linearity and range of positron. Silicon Photo Multipliers (SiPMs) are an option to readout the LYSO crystals in a strong magnetic field, hence opening the possibility of co-registration with MRI. 3. 05. 2007 Seminar SLAC 62

Compton Imaging 3. 05. 2007 Seminar SLAC 63

The main features of Compton Imaging are: The Mechanical collimator in the Anger Camera is replaced by Electronic Collimation. This removes the coupling between sensitivity and spatial resolution. This is achieved by having two detectors in coincidence: In the first detector the gamma rays are scattered by Compton Scattering on electrons in the detector material In the second detector the scattered gamma ray is absorbed 3. 05. 2007 Seminar SLAC 64

The measured quantities in Compton imaging are: x, y, z-co-ordinates in the first detector x, y, z-co-ordinates in the second detector Energy of recoil electron in first detector Energy of scattered photon in second detector Not measurable with Compton Cameras for medical applications: Direction of recoil electron, which leads to the conical ambiguity. This leads to more complicated image reconstruction algorithms. Expected improvements over Anger Camera: Factor ~5 in spatial resolution for probes Factor 5 to 50 improvement in sensitivity Due to Doppler effect smearing of the recoil energy resolution: Silicon is the only realistic semiconductor detector material for first detector 3. 05. 2007 Seminar SLAC 65

Results from a Demonstrator Test for a Compton Prostate Probe in 2005 3. 05. 2007 Seminar SLAC 66

Silicon detector and stack of 5 detectors 3. 05. 2007 Seminar SLAC 67

A Demonstrator set-up with stack of 5 Silicon pad sensor and 3 camera heads 3. 05. 2007 Seminar SLAC 68

Main Results Spatial resolution was measured for 4 energies; 57 Co (122 kev) and 133 Ba (272,302 and 356 kev). For the highest energy with a source-first detector distance of 11.3 cm: 5mm FWHM With a source Si distance of 3 cm this gives (simulation) 2-3 mm FWHM 3. 05. 2007 Seminar SLAC 69

Status: Spatial resolution in Silicon Demonstrated Next Demonstrator test foreseen before summer of 2007 with much improved camera head and improved silicon ( lower thresholds possible) 3. 05. 2007 Seminar SLAC 70

A High Resolution Small Animal PET Scanner based on Compton Scatter Events in Silicon Pad Detectors 3. 05. 2007 Seminar SLAC 71

A Very High Resolution PET Scanner for small animals based on Compton Scattering events is proposed: The Concept Three Major Coincidence Events BGO detector Si-Si : Very High Resolution BGO- BGO Si detector Si-Si Si-BGO Si-BGO : High Resolution BGO-BGO : Conventional PET Resolution 3. 05. 2007 Seminar SLAC 72

Simulation results with this configuration BGO ring Efficiency for different event classes Radial Posn. (mm) Detection Efficiency (%) Single Single Single BGO BGO - BGO 0 1.05 8.83 20.84 6 0.96 8.96 20.69 12 1.04 8.94 19.70 18 1.19 9.06 18.17 Calculated for point source in center plane. Only single scattering or absorption interactions in the silicon detector are included. Back scattered photons from BGO and events without full energy deposition are excluded. 3. 05. 2007 Seminar SLAC 73

Compton PET Test Bench Silicon detector BGO detector VATAGP3 HAMAMATSU PMT R2497 4.5 cm 2.2 cm and 1 mm thick 32 16 (512) pads, 1.4 mm 1.4 mm pixel size Energy Resolution 1.39 kev FWHM for Tc 99m 5.3 cm 5 cm and 3 cm thick 8 4 array, 12.5 mm 5.25 mm crystal size Energy Resolution 22% FWHM for Na-22 Harris Kagan Imaging 2006, June 26-30, Stockholm

Prototype PET Instrument Single-slice instrument using silicon and BGO Silicon detector Silicon detector Disassembled Assembled Harris Kagan Imaging 2006, June 26-30, Stockholm

3. 05. 2007 Seminar SLAC 76

Resolution Uniformity 5 4 3 2 1 0 0 1 2 3 4 5 cm Source pairs at 5, 10, 15, & 20mm off-axis Sinogram The sources in each pair are clearly separated at appropriate sinogram angles Harris Kagan Imaging 2006, June 26-30, Stockholm

Compton PET: Intrinsic Resolution F-18 0.254 mm 0.127 mm SS_steel wall Needle 25G (ID = 0.254 mm, OD = 0.5mm, SS_steel wall = 0.127 mm) 5 5 4 4 3 3 2 1 2 1 0 0 1 2 3 4 5 cm Harris Kagan Image Resolution = 700 μm FWHM 0 0 1 2 3 4 5 cm Imaging 2006, June 26-30, Stockholm

A Prototype Module for Small Animal X- Ray CT Fast Counting Chip with Energy Window Edgeless Micro Pad 1mm thick Silicon Detectors; In prototype module pitch 130 micron. Probably needs to be reduced. 3. 05. 2007 Seminar SLAC 79

A Prototype Module The complete module with four ASICs and two detectors, 512 pixels. Fixed on the alu base plate with 3 screws. Note that the 2 detectors are slightly wider than the PCB and the alu support allowing in principle to arrange several modules side by side ( two sided Butting ) with minimum distance between detectors 3. 05. 2007 Seminar SLAC 80

Photograph of middle of module 3. 05. 2007 Seminar SLAC 81

3. 05. 2007 Seminar SLAC 82

Conclusion The silicon sensor with a dominant role in medical imaging is still the photo diode array for readout of plastic scintillators in X-ray CT. Replacement of this technology might come in the form of Cd(Zn)Te readout with very high speed counting ASICs There are many attempts and projects to apply HEP developed technology,based on silicon detectors, in medical imaging and develop instruments for imaging and actually get those into hospitals. So far only one new device (to my knowledge), a digital mammography low dose (exposure) X-ray CT scanner (SECTRA [www.sectra.com]) based on silicon microstrip detectors, is on its way to become a standard in hospitals. Other promising applications are studied with intensive R&D efforts. Most important impact of silicon radiation sensors and submicron front-end electronics will be in PET and SPECT with impressive performance improvements of SiPM processing (my prediction). There might be a hard time coming up for Photo Multiplier tubes which dominate the readout concepts of present day nuclear imaging devices. 3. 05. 2007 Seminar SLAC 83