A1262227501KNtZp

1
Selected Advances in PET Instrumentation
Simon R. Cherry1, Yongfeng Yang1, Yibao Wu1,
Jinyi Qi1 and Ramsey Badawi2 1Department of
Biomedical Engineering 2Department of
Radiology University of California, Davis
2
With thanks to other members of the UC Davis
Nuclear Medicine Basic Science Team
Julie Sutcliffe (Biomedical Engineering)
Jinyi Qi (Biomedical Engineering)
Ramsey Badawi (Radiology)
3
and to Lab Members Past and Present

• Stefan Siegel Sara St. James Douglas Rowland
• Yiping Shao Purushottam Dokhale Stephen
  Rendig
• Guido Zavattini Freek Beekman Chris Griesemer
• Arion Chatziioannou Melissa Freedenberg
  Jennifer Fung
• Amy Moore Bernd Pichler
• Yuan-Chuan Tai Robert Silverman
• Niraj Doshi Hongjie Liang
• Randy Slates Ciprian Catana
• Andrew Goertzen Huini Du
• Daniel Rubins Yibao Wu
• Jennifer Stickel Martin Judenhofer
• Tim Tran Yongfeng Yang
• Gregory Mitchell Emilie Roncali
• Shrabani Sinha Julie Bec
• Ruby Gill Abhijit Chaudhari
• Changqing Li Andrew Bowery
• Erkan Mumcuoglu David Boucher

4
? Decay and Annihilation
5
Positron Emission Tomography

• PET quantitatively and non-destructively measures
  the 3-D distribution of radiolabeled biomolecules
  in vivo
• Primary tasks for imaging physics/engineering
• detect as many decays as possible
• put events in the right place
• make corrections and reconstruct quantitative
  images (Bq/cc)

6
Positron Emission Tomography

• Inject radiotracer
• Detect (scintillation detectors) two annihilation
  photons in coincidence
• Defines line along which annihilation lies
• Collect 107-108 events
• Use reconstruction algorithms to compute image of
  radiotracer distribution using all the different
  angular views
• Analyze data
• Lesion detection
• Quantify radiotracer distribution
• Tracer kinetics

7
Major Advances in Last 20 Years
8
Whole-Body Imaging
Guerrero TM, Hoffman EJ, Dahlbom M et al, IEEE
Trans Nucl Sci 37 676-680, 1990
Images courtesy of Magnus Dahlbom, UCLA
Impact Key concept that led to the clinical
adoption and reimbursement of PET
9
3-D PET
Colsher JG. Phys Med Biol 25 103-115, 1980.
Daube-Witherspoon M and Muehllehner G. J Nucl
Med 28 1717-24, 1987. Kinahan PE, Rogers JG.
IEEE Trans Nucl Sci 36 964-968, 1989. Townsend
DW, Geissbuhler A, Defise M et al. IEEE Trans
Med Imag 10 505-512 1991. Cherry SR, Dahlbom M,
Hoffman EJ. J Comput Assist Tomogr 15 655-668,
1991.
2-D
3-D
Impact Allowed gt5-fold improvements in
sensitivity with same instrument
10
Lutetium Oxyorthosilicate (LSO)and related
scintillators (MLS, LYSO)
Melcher CL and Schweitzer JS. Nucl Instr Meth
A314 212-214, 1992.
NaI(Tl) BGO LSO
Brightness (photons per MeV) 38000 8200 25000
Decay time (nsecs) 230 300 42
Mean free path _at_ 511 keV (cm) 2.88 1.05 1.16
Impact Improved spatial resolution, timing
resolution and count-rate performance
11
Iterative ReconstructionOSEM and System Modeling
Hudson HM and Larkin RS. IEEE Trans Med Imag 13
601-609, 1994. Qi JY, Leahy RM et al. IEEE Trans
Nucl Sci 45 1096-1103, 1998. Fessler JA. IEEE
Trans Med Imag 13 290-300, 1994.
FBP
iterative
Images courtesy of Richard Leahy, USC
FBP
iterative
Impact Dramatic improvement in image quality
for a wide range of PET studies
12
PET/CT
Beyer T, Townsend DW, Brun T et al. J Nucl Med
41 1369-1379, 2000.
Images courtesy of David Townsend, University of
Tennessee
Impact Synergy results in improved
sensitivity/specificity for lesion detection, and
CT used for PET attenuation correction
13
Small Animal Imaging
Bloomfield PM, Rajeswaran S, Spinks TJ et al.
Phys Med Biol 40 1105-1126, 1995. Lecomte R,
Cadorette J, Rodrigue S et al. IEEE Trans Nucl
Sci 43 1952-1957, 1996. Cherry SR, Shao Y,
Silverman RW et al. IEEE Trans Nucl Sci 44
1161-1166, 1997. Jeavons AP, Chandler RA, Dettmar
CAR. IEEE Trans Nucl Sci 46 468-473, 1999.
Images courtesy of Yongfeng Yang
Image courtesy of Hongjie Liang
Impact Opened up preclinical imaging and drug
development to nuclear medicine technologies,
ability to translate PET from mouse to man
14
Improvements in PET(last 20 years)

• gt 10-fold increase in volumetric spatial
  resolution (100-fold for preclinical imaging)
• gt10-fold increase in sensitivity
• Fast, whole-body imaging
• Correlated with anatomy (CT)
• Extension into preclinical research
• New targeted imaging probes

15
Recent Advances

• Time of Flight PET (for human imaging)
• Improving CNR by constraining annihilation
  position using timing information
• Depth-Encoding Detectors (for small animal
  imaging)
• The pathway to major improvements in sensitivity
  and spatial resolution for preclinical imaging
• Hybrid Imaging
• Breast PET/CT
• PET/MR

16
Time of Flight PET

• Speed of light is
• 3 x 108 m/s
• 30 cm/ns
• 5 mm spatial resolution would require 33 psec
  timing resolution!

17
Time of Flight Constraint
without time-of-flight
with time-of-flight
noise reduction
18
Timing Resolution of Scintillators
Material Coinc. LSO 221 ps LaBr330 Ce 165
ps CeBr3 198 ps LuI330 Ce 195 ps
Data courtesy of Bill Moses Marcus Ullisch Kanai
Shah
19
Time of Flight Gains

• Favorable scenario 40 cm diameter patient, 300
  ps timing resolution
• Effective increase in signal-to-noise x 3
• Less favorable scenario 30 cm diameter patient,
  600 ps timing resolution
• Effective increase in signal-to-noise x 1.8

20
Whole-Body FDG with TOF
non TOF
Philips Gemini TF
TOF
PET scanner LYSO 4 x 4 x 22 mm3 28,338
crystals, 420 PMTs 70-cm bore, 18-cm axial
FOV 600 ps timing resolution CT scanner
Brilliance 16-slice
Data courtesy of Joel Karp
21
Factors Determining Resolution and Sensitivity

• Spatial Resolution
• Crystal width
• Positron range
• Scanner geometry
• Non-colinearity
• Depth of interaction
• Detector scatter
• Sensitivity (?SNR2)
• Crystal thickness
• Scanner geometry
• Solid angle coverage

22
Effect of Detector Size in Small Animal PET
1 mm
best achievable resolution
detector size
23
Factors Determining Resolution and Sensitivity

• Spatial Resolution
• Crystal width
• Positron range
• Scanner geometry
• Non-colinearity
• Depth of interaction
• Detector scatter
• Sensitivity (?SNR2)
• Crystal thickness
• Scanner geometry
• Solid angle coverage

Image courtesy Bernd Pichler
24
Depth-Encoding PET Detectors
PSAPD 2 (8 x 8 mm2)
13 x 13 LSO array (0.5 x 0.5 x 20 mm3 elements)
PSAPD 1 (8 x 8 mm2)
25
Depth of Interaction Resolution
PSAPD
26
New Small Animal PET Scanner Design with
Depth-Encoding Detectors
27
Whole-Body PET/CT
Image courtesy of UCLA
28
Breast CT (bCT)
John Boone, Ph.D. Radiology/BME
29
Dedicated Breast PET

• Advantages
• Higher spatial resolution
• Higher sensitivity
• Should translate to better image quality
• Detection of smaller lesions?
• Disadvantages
• Limited to imaging breast
• Difficulty visualizing chest wall

30
Breast PET System
two detector heads 12 cm x 12 cm detectors 3 x
3 x 20 mm LSO crystals 1296 elements/detector st
ep shoot acquisition
Image courtesy Ramsey Badawi, Radiology, UC Davis
31
Integration with breast CT
Movie courtesy Ramsey Badawi, Radiology, UC Davis
32
Whole-body PET/CT
Dedicated breast PET/CT
Images courtesy of Spencer Bowen
49-yo woman, palpable mass in right upper outer
quadrant. Invasive mammary carcinoma confirmed
at biopsy.
33
Question

• The disadvantage of dedicated breast PET/CT is
• Examination results in higher radiation dose to
  the patient
• Lesions close to the chest wall are difficult to
  image
• It is not possible to use contrast for the CT
  component
• Imaging times are longer than for whole-body
  PET/CT

34
Answer

• The disadvantage of dedicated breast PET/CT is
• Examination results in higher radiation dose to
  the patient
• Lesions close to the chest wall are difficult to
  image
• It is not possible to use contrast for the CT
  component
• Imaging times are longer than for whole-body
  PET/CT

35
PET/MRI - motivation

• Like PET/CT provides
• Near-perfect registration of image data
• Anatomically-guided interpretation of PET or
  SPECT data
• Anatomic priors for reconstruction and data
  modeling
• Additional Advantages
• No additional radiation dose
• Can exploit soft-tissue contrast of MRI
• Can be combined with advanced MRI techniques such
  as fMRI, MRS, DWI and MR molecular imaging
  methods
• Additional Disadvantages
• Does not directly provide attenuation correction
  for PET
• Technically more difficult and likely more
  expensive

36
Technical Challenges in PET/MRI

• Interference on PET
• Static magnetic field
• Electromagnetic interference from RF and
  gradients
• Interference on MR
• Electromagnetic radiation from PET/SPECT
  electronics
• Maintaining magnetic field homogeneity
• Eddy currents
• Susceptibility artifacts
• General Challenges
• Space
• Environmental factors (temperature, vibration)
• Cost

37
Approaches to PET/MRI
tandem PET/MRI
integrated PET/MRI
PET
interference easier to avoid largely use
existing hardware least expensive
simultaneous PET/MRI possible higher
throughput best image registration
38
MR Compatible PET System
Concept
Animal MR System
PET Detectors
Magnet
Gradient Coils

RF coil
39
MR-Compatible PET Detector Module
optical fiber bundle
position-sensitive avalanche photodiode
preamplifiers
scintillator array
40
PET Insert
scintillator ring
optical fibers
PSAPDs
preamplifiers
41
PET-MRI Set Up
Number of modules Ring diameter Axial FOV Transaxial FOV Number of crystals Insert length Insert outer diameter 16 60 mm 12 mm 35 mm 1024 55 cm 11.8 cm
PET insert
RF coil
Gradient set
42
64Cu-albuleukin
In Vivo Simultaneous PET/MRI
30 mins
6 hrs
19 hrs
44 hrs
in collaboration with Dr. Andrew Raubitschek,
City of Hope
106 MC38-CEA cells, 3 mm tumors
43
(No Transcript)
44
FDG-PET guided MRS
Region Cho/Cr
High FDG Tumor 3.1
Low FDG Tumor 1.7
Muscle Negligible
PRESS TR/TE 1685/10ms 3mm3 voxel VAPOR H2O
suppression
High Choline may suggest high membrane turnover
rate cell proliferation
45
Next Generation PET/MR
46
Human MRI/PET
PET
Fused PET/MR
MRI
Courtesy of Ciprian Catana, Bruce Rosen and Greg
Sorenson, MGH/Harvard and Siemens
47
Other Developments

• Detector materials
• New scintillators (e.g. LuI3)
• Dense semiconductors (e.g. TlBr)
• New photodetectors (e.g. SiPMs)
• Fast, massively parallel electronics
• Larger axial field of view PET scanners
• Improved theoretical understanding of iterative
  algorithms, also faster algorithms
• Quantitative analysis, better use of time domain

48
Acknowledgments
R01 EB000561, R01 EB000230, R01 EB000993, R01
CA121783, U24 CA110804
Instrumentation Preclinical Molecular
Imaging Simon Cherry Yongfeng Yang Gregory
Mitchell Yibao Wu Changqing Li Shrabani
Sinha Sara St. James Huini Du Melissa
Freedenberg Bo Peng Computational Molecular
Imaging Jinyi Qi Guobao Wang Nannan Cao Lin
Fu Michel Tohme Jinxiu Liao Clinical Molecular
Imaging Physics Ramsey Badawi Abhijit
Chaudhari Jonathan Poon Spencer Bowen
Center for Molecular Genomic Imaging Douglas
Rowland David Kukis Stephen Rendig Chris
Griesemer Jennifer Fung David Boucher Lina
Planutyte Recent Lab Members Ciprian Catana
(MGH/Harvard) Bernd Pichler (U. Tübingen) Martin
Judenhofer (U. Tübingen) Jennifer Stickel Hongjie
Liang (Philips Medical) Purushottam Dokhale (RMD
Inc.) Recent Collaborators Richard Leahy
(USC) Joyita Dutta (USC) Sangtae Ahn (USC) Kanai
Shah (RMD Inc.) Richard Farrell (RMD
Inc.) Russell Jacobs (Cal Tech) Daniel Procissi
(CalTech) Thomas Ng (CalTech)