Outline
PET/CT Technology Updates,
Quality Assurance and Applications
• Evolution of PET instrumentation
• 2D 3D
• Dedicated
Osama Mawlawi, Ph.D.
PET/CT
• Advantages and short comes of PET/CT
Department of Imaging Physics
MD Anderson Cancer Center
• Possible solutions to the short comes
• QA for quantitative PET imaging
• New features in PET/CT scanners
• Future developments in PET/CT imaging
O. Mawlawi MDACC
O. Mawlawi MDACC
WHAT IS PET?
Principles of PET Imaging
• Functional imaging modality as compared to structural
• Injection of a radioactive tracer to image chemical/biological
processes.
• Radioactive tracer decays by Positron Emission.
• Functional images show:
• Blood flow
• Glucose metabolism
• Receptor density
• When the tracer is introduced into the body, its site-specific
uptake can be traced by means of the labeled atom.
O. Mawlawi MDACC
O. Mawlawi MDACC
Annihilation
Positron Decay
Detector ring
p = n + β+ + υ
• A nucleus with too low a neutron-to-proton
ratio converts a proton to a neutron,
emitting a positron (β+) and a neutrino (υ)
to carry off the excess energy.
511keV
Photon
LOR
electron
positron
Nucleus
511keV
Photon
β+
O. Mawlawi MDACC
E=mC2
Electron rest mass = 9.11x10-31 kg
C=3.0x108 m/s
1 joule =1 kg m2/s2
1 e = 1.602x10-19 coulomb
1 eV = 1.602x10-19 J
Thus: E =1.02 MeV
O. Mawlawi MDACC
Ax
ia
l
LOR Data Storage
Modified from http://www.crump.ucla.edu/software/lpp/lpphome.html
O. Mawlawi MDACC
O. Mawlawi MDACC
Ax
ia
l
Sample Sinograms
Modified from http://www.crump.ucla.edu/software/lpp/lpphome.html
O. Mawlawi MDACC
O. Mawlawi MDACC
Image reconstruction
Quantification: Power of PET
Measured Data
Random subtract
Normalize
Dead time
Correct Geometry
Calculate/subtract scatter
Correct Attenuation
FBP or IR reconstruction
Ideal measured data
after the application of several corrections
O. Mawlawi MDACC
From physics in Nuclear medicine Sorenson and phelps
O. Mawlawi MDACC
Overall LOR Determination process
x
x
y
y
Location
Timing
12 nsec
Energy
Location
Energy
Time
Location
Energy
Time
O. Mawlawi MDACC
O. Mawlawi MDACC
Physical Properties of PET Scintillators
8*8 Scintillation block Block / PMT setup
Material
BGO
GSO
LSO
NaI(Tl)
Density
7.1
6.7
7.4
3.7
LYSO*
-
Atomic number, Z
75
59
66
51
-
µ, cm -1 @ 511 keV
0.95
0.70
0.88
0.34
0.80
Mean free path, mm
10.5
14.1
11.4
29.1
12.5
Photofraction, %
40
25
32
17
35
Light Output,
photons/MeV
9000
8000
30000
41000
31500
Decay time, ns
300
60
40
230
40
λ, nm
480
440
420
410
-
Energy resolution,
%FWHM
12
9
10
8
15
Hygroscopic
No
No
No
Yes
No
Index of Ref.
2.15
1.85
1.82
1.85
-
Adapted from: Zanzonico et al. Sem. Nuc. Med. vol 34 No 2 2004
*Properties supplied by GE Healthcare.
O. Mawlawi MDACC
O. Mawlawi MDACC
2D vs 3D PET imaging
Type of recorded events
Scatter
Random
True
Total events = Trues + Randoms + Scatter
O. Mawlawi MDACC
From physics in Nuclear medicine Cherry Sorenson and phelps
Line source 2D in air
O. Mawlawi MDACC
Line source 2D in water
Inter-plane Septa
Line source 3D in water
O. Mawlawi MDACC
The theory and practice of 3D PET by Bendriem and Townsend
O. Mawlawi MDACC
Quantification: Power of PET
Measured Data
Random subtract
Normalize
Dead time
Correct Geometry
Calculate/subtract scatter
Correct Attenuation
From: Physics in Nuclear Medicine Cherry S et al.
Increase in sensitivity in 3D over 2D
* less injected activity
* less patient exposure
* shorter scanning time
P1 = e−µL1
FBP or IR reconstruction
Ideal measured data
less cost
O. Mawlawi MDACC
O. Mawlawi MDACC
Attenuation correction using
transmission scan
P2 = e−µL2
P1 * P2 = e−µL
Attenuation is dependant
on the path length and not
the depth of the source
of activity
Four ways to get an
attenuation map
1) Measured (MAC)
2) Calculated (CAC)
3) Segmented (SAC)
4) CT based (CTAC)
Nuclear Medicine: Diagnosis and therapy. Harbert J, Eckelman W., Neumann R.
O. Mawlawi MDACC
Transmission
rod source (68Ge)
O. Mawlawi MDACC
Dedicated PET Imaging
• Emission (2D mode)
• Transmission (scans are interleaved)
• 5 to 6 bed positions
• 8 min per position
• 5 EM, 3 Tx
• Total scan duration 50-60 min
Emission
15.5 cm
Tx rod sources
Transmission
Final
O. Mawlawi MDACC
O. Mawlawi MDACC
Disadvantages of dedicated PET
imaging techniques
• Transmission
– Noise due to low gamma ray flux from rod source
– Transmission is contaminated by emission data
• Scan duration
– Time consuming (emission & transmission )
– Increased patient movement (image blurring)
• Efficiency
– Decreased patient throughput
– Difficulty in correlating images to other diagnostic
modalities accurately
SUV = (Measured activity * Patient weight)/Injected dose
O. Mawlawi MDACC
O. Mawlawi MDACC
Hybrid Imaging
PET/CT: rationale
• Short duration, low noise CT-based attenuation correction
• Improved patient throughput (< 30 min total scan duration)
• Combines functional (PET) and anatomical (CT) imaging
GE Discovery ST
• PET and CT components can be operated independently
Siemens/CTI Biograph
• Fully quantitative, whole-body images for SUV calculation
• Improved patient comfort and convenience (single scan)
• Improved imaging accuracy for therapy planning & monitoring
Phillips Gemini
O. Mawlawi MDACC
O. Mawlawi MDACC
Short duration, low noise CT-based attenuation correction
Improved patient throughput
•
•
•
•
•
•
•
O. Mawlawi MDACC
Replace Tx scan by CT scan
3-5 min Tx scan duration
5-7 FOVs
15-35 min of Tx scanning
CT scan for whole body is about 1min
About 30 min of time saving per patient
Dedicated PET scan duration of 50 min
PET/CT scan duration of 20 min
O. Mawlawi MDACC
Combines functional
and
Anatomical data
CT based attenuation correction
Attenuation
Correction
Inherent
Image
Registration
O. Mawlawi MDACC
O. Mawlawi MDACC
Attenuation Scaling
with Energy
Attenuation Values are
Air
Energy Dependent
Muscle
•
Bone
0.3
•
Water (Soft Tissue) and
Bone are Different
µ/ρ
•
(cm/g)
0.2
Error
0.1
•
Solution is to Segment
Image into Tissue and
Bone and Scale
CT
PET
•
0
0
100
200
Energy
300
400
500
(keV)
For CT values < 0 ,
materials are assumed to
have an energy dependence
similar to water
For CT values > 0, material
is assumed to have an
energy dependence similar
to a mixture of bone and
water
The green line shows the
effect of using water scaling
for all materials
0.200
Attenuation at 511keV
•
Converting CT Numbers to
Attenuation Values
0.150
error
Bone/Water
0.100
0.050
0.000
-1000
Water/air
-500
0
500
1000
CT number measured at 140kVp
Separately
O. Mawlawi MDACC
O. Mawlawi MDACC
Converting CT Numbers to
Attenuation Values
•
PET/CT imaging artifacts are due to :
•
•
•
•
Contrast media
Truncation
Respiratory motion
Metal implants
•
•
O. Mawlawi MDACC
For CT values < 0 ,
materials are assumed to
have an energy
dependence similar to
water
For CT values > 0,
material is assumed to
have an energy
dependence similar to a
mixture of bone and
water
The green line shows the
effect of using water
scaling for all materials
0.200
Attenuation at 511keV
Types of Artifacts
0.150
error
Bone/Water
0.100
Water/air
0.050
0.000
-1000
-500
0
500
1000
CT number measured at 140kVp
O. Mawlawi MDACC
Transaxial
contrast bias
Coronal
Normal CT
Contrast enhanced CT
Contrast: Optiray 320, 100cc, 3cc/sec
O. Mawlawi MDACC
O. Mawlawi MDACC
Contrast Artifact- Intravenous
CT attenuation map scaling
Contrast
injected
0.200
Linear attenuation at 511keV 1/cm
0.180
140kVp non
contrast
120kVp non
contrast
100kVp non
contrast
140kVp BaSO4
0.160
0.140
0.120
0.100
120kVp BaSO4
0.080
100kVp BaSO4
0.060
140kVp I
0.040
120kVp I
100kVp I
0.020
0.000
-1000
-500
0
500
1000
1500
CT Hounsfield values
CT
O. Mawlawi MDACC
With contrast
PET CTAC
FUSED IMAGE
O. Mawlawi MDACC
Without contrast
SUV change
1.23 to 1.90
54% inc
61 year male with history of esophageal cancer
O. Mawlawi MDACC
O. Mawlawi MDACC
CT and PET Fields of View
X-ray focal spot
PET Detector
Truncation
PET 70cm image FOV
CT 50cm Fully sampled FOV
CT Detector
O. Mawlawi MDACC
O. Mawlawi MDACC
Truncation artifact appears as:
1. Rim of high activity concentration (AC) at the truncation
edge in the PET image
2. Area of low activity concentration in the region extending
beyond the CT FOV on the PET image due to the absence
of attenuation coefficients.
High AC in Rim
Low AC outside CT
FOV
Activity concentration recovered to within 5% of original value
O. Mawlawi MDACC
O. Mawlawi MDACC
CT
PET
FUSED
Wide View
Attenuation map
Truncated
Truncated
corrected
54 yrs. male with history of metastatic melanoma of
the skin. SUV changed from 3.25 to 6.05
O. Mawlawi MDACC
O. Mawlawi MDACC
Metal Artifacts
Artifact
Artifact
Metal Artifacts
O. Mawlawi MDACC
CT
PET CTAC
FUSEDIMAGE
O. Mawlawi MDACC
Breathing ArtifactCurvilinear Cold Areas
Photopenic
Artifact
Breathing Artifacts
CT
PET
FUSEDIMAGE
NON AC
• CT scans acquired at full inspiration result in a downward displacement
of the diaphragm due to the expansion of the lungs.
•Attenuation coefficients in this region will be underestimated since they
represent air rather than soft tissue.
• The attenuation corrected PET images will result in a curvilinear cold area
corresponding to this region.
O. Mawlawi MDACC
Breathing Artifact
O. Mawlawi MDACC
Breathing Artifact
Mismatch of lesion location between helical CT and PET
Mismatch of lesion location between helical CT and PET
O. Mawlawi MDACC
O. Mawlawi MDACC
Mismatch between PET and CT – Cardiac Application
Blurring due to object motion during data acquisition
Underestimation of Activity concentration
Courtesy of J. Brunetti, Holly name
Moving
Stationary
O. Mawlawi MDACC
O. Mawlawi MDACC
Average CT – 4D CT
X-ray on
off
Respiratory singal
Possible solutions
1st table position
(0 - 2cm)
2nd table position
(2 - 4 cm)
EI 0%
3rd table position
(4 - 6 cm)
EI 0%
EE 50%
EE 50%
CT images
4D-CT images
0
Helical CT
10
20
Helical CT
without thorax
30
40
50
60
70
80
90%
Combined helical and 4DCT averaging
respiration-averaged CT
1
Image averaging
Respiration-averaged CT
O. Mawlawi MDACC
Courtesy of Tinsu Pan, MDACC
2
O. Mawlawi MDACC
Clinical Studies
Mismatch 1:
CT diaphragm
position lower
than PET
+57%
CT
PET
Fused
RACT
PET
(A)
Fused
No AC
(B)
Mismatch 2:
CT diaphragm
position higher
than PET
(C)
Courtesy of Tinsu Pan, MDACC
O. Mawlawi MDACC
Improve the restaging after chemo
(D)
Courtesy of Tinsu Pan, MDACC
O. Mawlawi MDACC
Impact on treatment planning
Old GTV
Misregistration lead to a false positive response to chemo.
New GTV
A true negative response when misregistration is removed with ACT
Courtesy of Tinsu Pan,
MDACC
O. Mawlawi
MDACC
Previous GTV was outlined based on CT and clinical PET without motion correction.
New GTV was redefined based on the correct information from PET with ACT.
Courtesy of Tinsu Pan, MDACC
O. Mawlawi MDACC
Gated PET (Used to image repetitively moving objects: cardiac,
respiratory)
Trigger
4D-Gated PET
Trigger
1
8
2
1
3
4
8
7
2
7
6
3
6
4
5
5
time
Bin 1
Bin 8
• Prospective fixed forward time binning
• Ability to reject cycles (cardiac) that don’t match
• Single 15 cm FOV Gated PET
• User defined number of bins and bin duration
• As number of bins increase, the duration and motion per bin decreases.
However images will be noisy unless acquired for longer durations.
O. Mawlawi MDACC
O. Mawlawi MDACC
Phantom Study
Impact of Whole-body Respiratory Gated PET/CT
Coronal
Sagital
Mip
Axial
Sphere (volume)
1 (16.5 cc)
Gate d
31624
Static
33546
Error (%)
-6.1
2 (8.4 cc)
31316
21872
30.2
3 (4 cc)
4 (2 cc)
5 (1 cc)
6 (0.5 cc)
29919
22857
19087
11897
20607
13643
8264
5143
31.1
40.3
56.7
56.8
Static wholebody
Single respiratory phase
(1 of 7, so noisier)
1 cc lesion on CT
• The max SUV of the lesion goes from 2 in the static image to 6 in the
respiratory-gated image sequence
Courtesy of P Kinahan, UW
O. Mawlawi MDACC
New CT Application… Advantage
4D CT
4D-PET/CT
Respiratory tracking with Varian RPM optical monitor
CT images acquired over complete respiratory cycle
“Image acquired”
signal to RPM
system
X-ray on
First couch position
Second couch position
Third couch position
Respiratory motion defined retrospective gating
O. Mawlawi MDACC
O. Mawlawi MDACC
4D PET/CT
Phantom Study
2 cm
18FDG
Sphere-to-Background Ratio: 5.18
Total Activity: 2.49mCi
Scan Duration: 3 min
Number of Time Bins: 10
O. Mawlawi MDACC
O. Mawlawi MDACC
Phantom Study Results
Quality Assurance in Quantitative
PET/CT
1200
1200
1200
1200
1200
1000
1000
1000
1000
1000
800
800
800
800
800
600
600
600
600
600
400
400
400
400
200
200
200
200
0
0
0
0
5 10 15 20 25 30 35 40 45
Reference
5 10 15 20 25 30 35 40 45
gated
5 10 15 20 25 30 35 40 45
static
400
200
0
5 10 15 20 25 30 35 40 45
MIR
(real
motion)
5 10 15 20 25 30 35 40 45
MIR
(estimated
motion)
O. Mawlawi MDACC
O. Mawlawi MDACC
Definitions
Quantitative PET Performance
• Quality Assurance
– “….. systematic monitoring and evaluation of the various
aspects of a project, service, or facility to ensure that
standards of quality are being met”
–
–
•
http://www.webster.com/dictionary/quality%20assurance
•
3 Factors that may affect SUV measurements
•
Patient Compliance
–
–
–
–
Quality Control
– “an aggregate of activities (as design analysis and inspection for defects)
designed to ensure adequate quality especially in manufactured
products”
•
–
–
–
–
http://www.webster.com/dictionary/quality%20control
O. Mawlawi MDACC
Fasting
Blood Glucose levels
Scan Conditions
Scan time post injection
Patient anxiety/comfort during uptake (room temperature, etc.)
Patient motion during acquisition
PET/CT artifacts
Intrinsic System Operating Parameters
Calibration
QA – Maintenance of operating parameters
System performance characterization
Image processing algorithms
O. Mawlawi MDACC
PET Detector Array Format and
Design
PET Detector Calibration Overview
Installation/
Replacement of
Detector block, circuit
Board, etc.
Discovery STe example…
• Detector Array consists of 35 modules (0-34)
• Each module consists of 8 blocks (0-7)
• Each block consists of 8 x 6 BGO crystal array
• 1 Timing circuit per crystal pair
• 35*8*(8x6)= 13440 total detectors
0 1 2 3 4
RMS < x
Update Gains
RMS > x
Update Position
Map
Well Counter
Correction
(Q)
Geometric
Calibration
(Q)
………………………………………………………………………………………………………….
Coincident
Calibrations
Energy Update
0,1
12 hour Norm
(Q)
2,3
CTC
4,5
Singles
Calibrations
6,7
DQA Check
Set the Baseline
O. Mawlawi MDACC
Daily Quality Assurance Baseline
• A baseline QA acquisition is taken after Singles and
Coincidence calibration are performed following
installation of the system
• A baseline QA acquisition is also performed
whenever a detector block or circuit board has
been replaced
O. Mawlawi MDACC
PET Daily QA Scan
– Uses Ge-68 source pin that rotates across face of detector
array
– Includes a ~2 min normalization component in which
counts from the source flood are acquired
– Also includes a daily CTC after flood acquisition is
complete
• Used to construct ratio image comparing current
DQA results with this baseline acquisition
– Generates a visual map of detector array used to validate
detector functionality
– Generates statistics used to indicate scanner performance
O. Mawlawi MDACC
O. Mawlawi MDACC
PET Daily QA Scan
Daily Quality Assurance Scans
• Purpose
– To ensure that optimal clinical images will be obtained
from the scanner on a daily basis
– To quantitatively compare scan results with manufacturer
and institutional specifications and standards
– To enforce appropriate device usage by the technologist
– To allow baseline data to be available in order to derive
functional trends in image quality
O. Mawlawi MDACC
O. Mawlawi MDACC
Pre-Normalization
PET Daily QA Scan
Post-Normalization
Normalization
sinogram
O. Mawlawi MDACC
Courtesy of magnus Dahlbom
O. Mawlawi MDACC
Coincidence Calibrations
Coincidence Calibrations
• Well Counter Correction (WCC)
•
Pre-calibrated Phantom
O. Mawlawi MDACC
O. Mawlawi MDACC
Normalization effects
Coincidence Calibrations
• Pre-calibrated Phantom
– Pros
• Eliminates need to prepare a source, fill and calibrate a phantom
– Cons
• Based on phantom manufacturer’s dose calibrator readings
• Patient studies based on site dose calibrator readings
• Therefore, there exists a potential discrepancy between the dose
calibrators used to calibrate the scanner, and the dose calibrators
used to assay the doses of patient studies performed on the scanner
(un-correlated)
Good Norm
O. Mawlawi MDACC
Bad Norm
O. Mawlawi MDACC
Effect of bad Normalization
Effect of wrong WCC
SUVmax=4.5
Increased
activity due to
error in
normalization
O. Mawlawi MDACC
Effects of Bad Blocks
O. Mawlawi MDACC
Effects of Bad Blocks
O. Mawlawi MDACC
O. Mawlawi MDACC
Effects of Bad Blocks
Normalized to Baseline for each Sphere
105
sphere
37 mm
100
sphere
28 mm
95
percent
90
sphere
22 mm
85
sphere
17 mm
80
75
sphere
13 mm
70
sphere
10 mm
5
4,
,5
k4
6b
lk
22
bl
M
M
4,
5
M
5b
lk
4,
5
4,5
k4
9b
lk
22
bl
M
M
4,
5
M
5b
lk
5b
lk
4,
5
5b
lk
M
4
k2
,3
4
6b
lk
M
22
bl
5b
lk
4,
5
M
4,
5
M
M
5
4,
9b
lk
M
4,
5
5b
lk
M
M
5b
lk
5
3
2,
5b
lk
5b
lk
M
4,
5
5b
lk
5b
lk
M
M
M
4
lk
5b
lk
22
b
M
M
5
ba
se
li n
e
fi n
al
ba
se
l in
e
65
O. Mawlawi MDACC
Quantitative PET Performance
O. Mawlawi MDACC
Effect of Reconstruction Algorithm
a
c
e
b
d
f
3 Factors that may affect SUV measurements
•
Patient Compliance
–
–
•
–
–
–
–
•
–
–
–
–
Fasting
Blood Glucose levels
Scan Conditions
Scan time post injection
Patient anxiety/comfort during uptake (room temperature, etc.)
Patient motion during acquisition
PET/CT artifacts
Intrinsic System Operating Parameters
(a) 2D-FBP; (b) 2D-OSEM; (c) 3D-RP; (d) 3D-IR; (e) 3D-FORE+FBP; (f) 3D-FORE+OSEM.
AC max under different reconstruction algorithms (Bq/cc)
Calibration
QA – Maintenance of operating parameters
System performance characterization
Image processing algorithms
(x104)
O. Mawlawi MDACC
Sphere 1
Sphere 2
Sphere 3
Sphere 4
Sphere 5
Sphere 6
Algorithm a
3.03
2.74
2.52
2.30
1.75
Algorithm b
3.14
3.14
2.82
2.90
2.14
1.74
Algorithm c
2.83
2.76
2.62
2.41
1.60
1.07
1.12
Algorithm d
2.93
2.92
2.98
3.21
2.19
1.33
Algorithm e
2.78
2.75
2.64
2.40
1.59
1.10
Algorithm f
2.85
2.74
2.71
2.52
1.65
1.09
O. Mawlawi MDACC
A
B
3D IR
2it 20sub
3D FORE
2it 5sub
C
D
3D FORE
5it 28sub
3D FORE
FBP
Max SUVbw
in different
spheres
1
2
3
4
5
6
10.8
10.3
9.9
8.9
6.5
3.8
B
8.6
7.6
6.6
5.5
3.9
2.8
C
10.3
9.7
9.1
8.3
6.3
4.7
D
10.4
9.3
9.3
7.7
5.9
4.8
A
O. Mawlawi MDACC
Effect of scan duration and count density
on maximum AC
2D
4
x 10
Max d37
12
10
8
6
x 10
scan1
scan2
scan3
scan4
8
7
6
5
4
3
4
2
2
0
0
Effect of smoothing on measured AC
3D
4
9
scan1
scan2
scan3
scan4
14
Max d37
16
O. Mawlawi MDACC
20
40
60
80
100
120
Duration(sec)
140
160
180
1
0
20
40
60
80
100
120
140
160
180
Duration(sec)
O. Mawlawi MDACC
O. Mawlawi MDACC
Effect of ROI type
New PET Scanner Features
Courtesy ofO.Kinahan
UW
Mawlawi MDACC
O. Mawlawi MDACC
LIST Mode
PET data acquisition of different duration per FOV
X2
Y1
• Time ticks are fixed at 1msec intervals
• The number of events between time ticks depends
Static
3 min
on the amount of activity in the field of view
Y3
Y4
• The more activity, the more the events between
time ticks.
• Very flexible, data can be rebinned as static,
Static
3 min
dynamic, or gated.
• Requires large amount of memory
X1
X4
Static
5 min
X3
X1
Static
6 min
Y1
X2
Y2
TIME
X3
Y3
X4
1msec
X8
O. Mawlawi MDACC
Y8
X9
Y4
X5
Y5
X6
Y6
Y2
TIME
X7
Y7
1msec
Y9
X10
Y10
X11
Y11
O. Mawlawi MDACC
LIST data format
Tick marks
Advantages of List mode acquisition:
• Ability to post process the data using different prescriptions.
• Mode of acquisition can be varied (static, dynamic, gated….)
1 msec
• Duration of the scan can be changed (compare different scan durations)
• Chop off data segments based on user definition.
Section of a list file (141MB) using a 25MBq (0.7mCi) GeGe-68 rod source acquired for
a duration of 15 seconds.
O. Mawlawi MDACC
O. Mawlawi MDACC
LIST Mode
X1
Y1
X2
Y2
TIME
X3
Y3
1msec
X8
Y8
X9
X4
Y4
X5
Y5
X6
Y6
TIME
X7
Y7
1msec
Y9
X10
Y10
X11
Y11 TRIG
X12
Y12 TIME
List mode allows rebinning of acquired data into different durations
O. Mawlawi MDACC
O. Mawlawi MDACC
LIST data format
Triggers allows for double gating – resp and cardiac
Tick mark
Trigger input
New PET Scanner Designs
Section of a list file (141MB) using a 25MBq (0.7mCi) GeGe-68 rod source acquired for
a duration of 15 seconds.
O. Mawlawi MDACC
Improved sensitivity with TrueV
SIEMENS
O. Mawlawi MDACC
Advantages of the extended Axial Field-of-View
phototubes
• thicker LSO crystals
20 mm
scintillator
B
30 mm
• LSO volume increase: 50%
• sensitivity increase:
A
D
C
γ (511 keV)
40%
Sensitivity
planar sensitivity
+ 192 PMTs
• extended axial FOV
16.2 cm
FOV
21.6 cm
• increased sensitivity = shorter imaging per bed (or more counts)
• LSO volume increase: 30%
• sensitivity increase:
volume sensitivity
• larger axial FOV = fewer bed positions for same axial coverage
77%
3D (no septa)
Courtesy of D
Townsend
O. Mawlawi
MDACC
• reduction in imaging time (or dose) of ~2 for comparable quality
Courtesy of D
Townsend
O. Mawlawi
MDACC
Staging head and neck cancer Biograph 6 TrueV
Biograph TruePoint performance
Molecular
Imaging
Program
Patient NECRs
Noise Equivalent Count Rate
161 kcps @ 31 kBq/ml
NECR (cps)
TrueV
96 kcps @ 34 kBq/ml
Biograph TruePoint
SUV =
18.6
Activity (kBq/ml)
Sensitivity (cps/kBq)
Spatial resolution (mm)
Scan duration: 10 min
Scatter (%) (425 keV)
transverse axial
Biograph TP
4.49
TrueV
8.10
Biograph TP
4.1
4.5
True V
4.2
4.7
Biograph TP
34 %
TrueV
34 %
5/06
10.4 mCi, 90 min pi
108 lbs; 2 min/bed, 5 beds
4i / 8s; 5f
Courtesy of D
Townsend
O. Mawlawi
MDACC
Courtesy of D Townsend
BrainPET for MAGNETOM Trio,
The first results
BrainPET* for MAGNETOM Trio, A Tim System
in Progress. The information about this product is preliminary. The product is under development and is not commercially
available in the U.S., and its future availability cannot be ensured.
* Works
First ever in-vivo images simultaneously acquired by MR and PET (17.11.06)
PE
TY
TO
O
PR
* Works in Progress. The information about this product is preliminary. The product is under development and
is not commercially available in the U.S., and its future availability cannot be ensured.
115
Siemens Medical Solutions
Innovation is in our genes.
Molecular Imaging
MR
MR-PET
Courtesy of Townsend and Nahmias University of Tennessee, USA and
116 Pichler, Claussen University of Tuebingen, Germany Siemens Medical Solutions
Schlemmer,
PET
Innovation is in our genes.
Molecular Imaging
VUE Point – Iterative Reconstruction
BrainPET for MAGNETOM Trio, A Tim System
The first results
Conventional
Iterative Reconstruction
Randoms
Correction
Artifact free imaging even with the most demanding applications
Deadtime/
Normalization
Correction
Radial
Repositioning
Radial
Repositioning
Randoms
Correction
Quadrant
Scatter
Correction
Attenuation
Correction
VUE Point
Deadtime/
Normalization
Correction
Iterative Reconstruction w/ Distance Driven Projectors
Quadrant
Attenuation
Scatter
Correction
Correction
VUE Point HD (mid-2007 update)
T2 TSE
PET
MR-PET
ADC
Deadtime/
Normalization
Correction
Diffusion EPI sequence applied during PET acquisition.
Siemens Medical Solutions
117
Iterative Reconstruction w/ Distance Driven Projectors
Detector
Randoms
Volume Scatter
Geometry
Correction
Correction
Modeling
Attenuation
Correction
Innovation is in our genes.
Molecular Imaging
GE Healthcare, Molecular Imaging
Page 118
Courtesy of University of Tennessee, USA and University of Tuebingen, Germany
VUE Point - Volume DScatter Correction
1
VUE Point – Detector Geometry Modeling
D1
3.5
Native Geometry
Radial Repositoned
D1
Sinogram Bin Width (mm)
3
LOR
LOR
S
LOR
S
S
Discovery ST
2
1.5
1
-500
D2
D2
Quadrant Scatter Correction
• Direct block scatter LOR
• Single-quadrant processing
• High activity ratio artifact
2.5
D2
Volume Scatter Correction
Cross block scatter LOR
Multi-quadrant processing
Reduced artifact potential
-400
Radial Repositioning (former)
• Data into evenly spaced grid for Fourier
filtering
• Crystal position model includes effects of
block gaps and “average” penetration
• Output data determined by linear
interpolation between data points
-300
-200
-100
0
100
Radial Position (mm)
200
300
400
500
Detector Geometry Modeling
• Allows all corrections to be included
in loop (normalization/deadtime)
• Improvement in resolution
• Eliminates noise correlation
introduced by radial repositioning
• Reduction in spoke artifact
VUE Point – Targeted Iterative Reconstruction
VUE Point – Targeted Iterative Reconstruction
An Improved Solution: Keyhole Reconstruction
200
A more efficient method to accomplish the
targeted reconstruction can be described
in the following steps:
180
160
140
120
1
100
1.) Reconstruct the full SFOV at a pixel
pitch greater than the desired final pixel
pitch in the DFOV
80
60
40
20
0
-200
-150
-100
-50
0
50
100
150
2
200
2.) Zero out the pixels in this image
corresponding to the DFOV
• Targeted Iterative Reconstruction may lead to artifacts and
quantitative inaccuracy if the data outside the FOV is not considered
3
3.) Forward-project the resulting image to
create a sinogram representing the
activity outside the DFOV
4
4.) Subtract sinogram from the original
data set, to leave a data set representing
only activity from inside the DFOV
• A full FOV (left) and targeted FOV (center) reconstruction of a
simulated phantom (blue circle indicates the bounds of the targeted
region)
5
5.) Reconstruct this sinogram at the
desired pixel pitch to create the final,
targeted image
• Horizontal profiles through the full FOV (blue) and targeted FOV
(black) are shown in the graph on the right.
GE Healthcare, Molecular Imaging
GE Healthcare, Molecular Imaging
Page 121
Page 122
VUE Point – Targeted Iterative Reconstruction
VUE Point – Detector Normalization
• Detector Geometry Reconstruction enables normalization & deadtime to be
included in the iterative reconstruction loop
Rb82 Cardiac Exam
• Normalization & deadtime corrections in the loop increase detector variation stability
• Phantom simulation, 300k counts, 2D image
– edge crystals with 100%, 80%, 60%, 40%, 20% efficiency of the center crystals
Pre-correction
Normalization
Targeted Iterative
Reconstruction
“Keyhole Technique”
Targeted Iterative
Reconstruction
“Conventional Technique”
“In-the-loop”
Normalization
0
20
40
60
80
% degradation in efficiency of edge crystals w.r.t central crystals
GE Healthcare, Molecular Imaging
GE Healthcare, Molecular Imaging
Page 123
Page 124
Time-of-Flight Acquisition:
Principles
Motion Freeze
Impact of Whole-body Respiratory Gated PET/CT
Static wholebody
Single respiratory phase
(1 of 7, so noisier)
∆x =
1 cc lesion on CT
∆t
c
2
Back projection
along the LOR
C = 3*108 m/s
• The max SUV of the lesion goes from 2 in the static image to 6 in the
respiratory-gated image sequence
Courtesy of P Kinahan, UW
O. Mawlawi MDACC
Time of Flight and SNR, examples…
Colon cancer
∆t
∆x = c
2
SNRTOF ≅
Time
Resolution
(ns)
∆x
D
⋅ SNRconv
∆x
(cm)
SNR
improvement
(20 cm object)
SNR
improvement
(40 cm object)
0.1
1.5
3.7
5.2
0.3
4.5
2.1
3.0
0.5
7.5
1.6
2.3
1.2
18.0
1.1
1.5
Best measurement on LSO small crystal pair
CT
119 kg (BMI = 46.5) 30 min scan
non-TOF
TOF
MIP
[W. Moses]
Overall present time resolution on HiRez scanner
O. Mawlawi MDACC
Improvement in lesion detectability with TOF
non-Hodgkin’s lymphoma 136 kg (45 BMI)
Average-weight patient study
nonTOF 30 min
TOF
30 min
TOF
10 min
Tongue CA,
lung mets
67 kg
BMI = 29.0
CT
20 min scan
non-TOF
non-TOF
TOF
TOF
TOF tumor contrast (SUV) higher than non-TOF by 1.5
Improvement in lesion contrast with TOF
Courtesy of J. karp
Courtesy of J. karp
TOF tumor contrast superior to non-TOF for 10 min as well as 30 min scan
Mixed acquisition protocol
Static
FUTURE
Gated
Static
O. Mawlawi MDACC
O. Mawlawi MDACC
Effect on Partial voluming
13
All spheres contain
the same activity
concentration
17
22
10
Profile (10 mm)
28
37
Recovery (%)
Depth of interaction
100
22
28
37
17
50
13
0
10
Standard
8 x 8 detector
Sphere diameter
HI-REZ
13 x 13 detector
Recovery coefficients
Courtesy of Siemens
From physics in Nuclear medicine Cherry, Sorenson and phelps
O. Mawlawi MDACC
O. Mawlawi MDACC
Quantitative PET Performance
3 Factors that may affect SUV measurements
•
Patient Compliance
–
–
•
Fasting
Blood Glucose levels
Scan Conditions
–
–
–
–
•
Thank You
Scan time post injection
Patient anxiety/comfort during uptake (room temperature, etc.)
Patient motion during acquisition
PET/CT artifacts
Intrinsic System Operating Parameters
–
–
–
–
Calibration
QA – Maintenance of operating parameters
System performance characterization
Image processing algorithms
O. Mawlawi MDACC
O. Mawlawi MDACC
All of the following are types of events that a
PET scanner detects except:
25%
25%
25%
25%
1.
2.
3.
4.
Type of recorded events
True Events
Scatter Events
False Events
Random Events
Scatter
Random
True
Total events = Trues + Randoms + Scatter
10
Attenuation in PET imaging is
dependent on:
25%
P1 = e−µL1
25%
1. Annihilation photons path length
2. Depth of annihilation event in the
radioactive source
3. Amount of radioactivity present in the
source
25%
4. Number of detectors in the scanner
25%
10
P2 = e−µL2
P1 * P2 = e−µL
Attenuation is dependant
on the path length and not
the depth of the source
of activity
Four ways to get an
attenuation map
1) Measured (MAC)
2) Calculated (CAC)
3) Segmented (SAC)
4) CT based (CTAC)
Nuclear Medicine: Diagnosis and therapy. Harbert J, EckelmanO.W.,
Neumann
R.
Mawlawi
MDACC
All of the following are advantages of PET/CT
imaging over dedicated PET imaging except:
25%
25%
25%
25%
1.
2.
3.
4.
Shorter total scan duration
Better image resolution
Low noise attenuation correction
Automatic fusion of functional and
anatomical images
Disadvantages of dedicated PET
imaging techniques
• Transmission
– Noise due to low gamma ray flux from rod source
– Transmission is contaminated by emission data
• Scan duration
– Time consuming (emission & transmission )
– Increased patient movement (image blurring)
• Efficiency
– Decreased patient throughput
– Difficulty in correlating images to other diagnostic
modalities accurately
10
O. Mawlawi MDACC
Truncation artifact in PET/CT
imaging:
25%
25%
25%
Truncation artifact appears as:
1. Reduces activity concentration in the affected
area
2. Increases the activity concentration in the
affected area
3. Has no effect on activity concentration in the
area of interest
1. Rim of high activity concentration (AC) at the truncation
edge in the PET image
2. Area of low activity concentration in the region extending
beyond the CT FOV on the PET image due to the absence
of attenuation coefficients.
High AC in Rim
25% 4. Affects mainly small size patients
Low AC outside CT
FOV
10
O. Mawlawi MDACC
All of the following are PET/CT artifacts due
to the use of CT for attenuation correction of
PET data except:
25%
25%
25%
25%
1.
2.
3.
4.
Contrast media artifact
Tube current artifact
Metallic artifact
Breathing motion artifact
Activity concentration recovered to within 5% of original value
10
O. Mawlawi MDACC
A well counter calibration in PET
imaging is used to:
Types of Artifacts
PET/CT imaging artifacts are due to :
25% 1. Correct for variations in image uniformity
•
•
•
•
Contrast media
Truncation
Respiratory motion
Metal implants
25% 2. Correct for variations in detector gains
for differences in detector coincidence
25% 3. Correct
timing
detected count rate to activity
25% 4. Transform
concentration
10
O. Mawlawi MDACC
Coincidence Calibrations
• Well Counter Correction (WCC)
O. Mawlawi MDACC