Outline
Advances in PET Technology
New Crystals and Detector Designs
Frederic H. Fahey DSc
Children’s Hospital Boston
Harvard Medical School
frederic.fahey@childrens.harvard.edu
Positron Emission
•
•
•
•
•
•
Basics of PET Scanner Design
Data Acquisition
Review of Scintillation Materials
PET/CT
Time-of-Flight (TOF) PET
MicroPET
Detector Ring
18F
+
511 keV
e511 keV
1
Detector Blocks (GE Advance NXi)
True, Scatter and Random
Coincidence Detections
True
Scatter
Random
(R = 2 τ N1 N2)
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
Span of 3 Michelogram
17
*
*
*
*
*
*
*
18
* *
* * *
* * *
* * *
* * *
* * *
* * *
* * *
* * *
* * *
* * *
* * *
* * *
* * *
* * *
* * *
* * *
* *
*
*
*
*
*
*
*
17
*
*
*
*
*
*
*
16
*
*
*
*
*
*
*
15
*
*
*
*
*
*
*
14
*
*
*
*
*
*
*
13
*
*
*
*
*
*
*
12
*
*
*
*
*
*
*
11
*
*
*
*
*
*
*
10
*
*
*
*
*
*
*
9
*
*
*
*
*
*
*
8
*
*
*
*
*
*
*
7
*
*
*
*
*
*
6
*
*
*
*
*
5
4
3
2
1
*
*
*
*
1
2
3
4
5
6
7
8
9
10
11
12
13
14
*
*
*
*
*
*
15
*
*
*
*
*
16
*
*
*
*
17
18
18
Span of 7 Michelogram
2
2D Detector
Scintillator
Crystals
PMTs
End
Shields
Acquisition Modes
Septa
Axial Direction
2D
511 kev photon
3D
septa
are
removed
Courtesy of M. Graham, M. Madsen, U Iowa
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
1
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
2
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
3
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
4
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
5
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
6
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
7
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
8
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
9
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
10
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
11
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
12
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
13
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
14
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
15
GE 3D Projection view and Michelogram
11
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
18
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
17
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
16
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
15
*
*
*
*
*
*
*
*
*
*
*
*
*
*
14
*
*
*
*
*
*
*
*
*
*
*
*
*
13
Segment 2
16
Segment 1
Segment 3
*
*
10 *
9 *
8 *
7 *
6 *
5 *
4 *
3 *
2 *
1 *
12
11
1
2
3
4
5
6
7
7
9
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
8
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
9
10
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
11
5
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
12
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
13
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
14
3
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
15
1
*
*
*
*
*
*
*
*
*
*
*
*
*
*
16
*
*
*
*
*
*
*
*
*
*
*
*
*
17
*
*
*
*
*
*
*
*
*
*
*
*
2
4
6
8
10
18
3
3D PET
PET Slice Sensitivity
20
Slice Sensitivity (Rel. Units)
18
16
14
2D Span of 3
12
2D Span of 7
10
3D Ring Diff of 11
8
3D Ring Diff of 18
6
4
2
0
0
10
20
30
Slice Num ber
Criteria for Scintillation Material
•
•
•
•
•
Sensitivity drops off towards edges
4-5X increased sensitivity overall
Increased scatter (15% to 40%)
Increased randoms from out-of-field activity
Rebinning algorithms to apply 2D
reconstruction
• Some devices can acquire in 2D or 3D whereas
some can only acquire in 3D
• 3D in Brain, 2D (or 3D) in Whole Body
Crystal Identification
• Detection Efficiency (Stopping Power)
– High Effective Z
– High Density
• Light Output
– Good energy resolution
– Good crystal identification
• Decay Time
– Reduction of random coincidences
– Time-of-Flight PET
4
New Detector Materials
SCINTILLATOR
Rel. Light Output
NaI(Tl)
100
BGO
LSO
PET/CT
GSO
15-20
75
20-25
Peak Wavelength (nm) 410
480
420
440
Decay Constant (ns)
230
300
12,42
30-60
Density (g/mL)
3.67
7.13
7.40
6.71
Effective Z
51
75
66
59
Index of Refraction
1.85
2.15
1.82
1.85
Hygroscopic ?
Yes
No
No
No
GE Discovery ST PET/CT
• State-of-the-art PET combined with
state-of-the-art CT (up to 64 slice)
• Anatomical correlation
• CT-based attenuation correction
PET Attenuation Correction
P2 = e-µ(L-x)
X
P1 = e-µx
L
PTOT = P1 x P2
= e-µL
CT
PET
5
PET-CT Attenuation Correction
PET-CT Attenuation Correction
PET-CT Attenuation Correction
Dose from CT of PET-CT
CTADIvol (10 ye ar old, 0.8 s , 1.5:1 pitch)
Acquire CT Scan and reconstruct
Apply energy transformation
Reproject to generate correction matrix
Smooth to resolution of PET
Apply during reconstruction
2500.00
CTADIvol (mrad)
•
•
•
•
•
2104 mrad
2000.00
80 kVp
1500.00
100 kVp
120 kVp
1000.00
140 kVp
500.00
0.00
0
30 mrad
50
100
150
200
Tube Curre nt (m A)
6
Dose from CT of PET-CT
Quality of CTAC
CTADIvol (160 m A, 0.8 s , 1.5:1 pitch)
3000.00
CTADIvol (mrad)
2500.00
New Born
2000.00
1 Y ear Old
5 Y ear Old
1500.00
10 Y ear Old
1000.00
Med Adult
80 kVp
10 mA
0.5 s/rot
1.5:1
140 kVp
160 mA
0.8 s/rot
1.5:1
500.00
0.00
70
90
110
130
150
Tube V oltage (k V p)
ED from
14 mCi of FDG
1 rad
Effect of Patient Size
80 kVp, 10 mA, 0.5 s/rotation
Quality of CTAC
Accuracy of Atte nuation Correction w ith Patie nt Size
0.0920
80 kVp
10 mA
0.5 s/rot
1.5:1
140 kVp
160 mA
0.8 s/rot
1.5:1
Linear Attenuation Coefficient (cm-1)
0.0900
New born
0.0880
1 YO
5 YO
0.0860
10 Y O
15 Y O
0.0840
Small A dult
Med Adult
0.0820
Large A dult
0.0800
0.0780
80/10/.5
7
Effect of Patient Size
0.092
0.09
Large A dult
0.088
Medium Adult
0.086
Small A dult
0.084
0.082
•
•
•
•
Siemens Biograph
GE Discovery ST
GE Discovery STE
Philips Gemini
14
0/
10
/0
.8
.5
.8
/0
14
0/
10
/0
0/
10
12
12
0/
10
/0
.5
.8
.5
/0
/0
0/
10
10
0/
0/
10
10
80
/1
0/
0.
5
0.
8
0.08
80
/1
Linear Atten Coef for 511 keV (cm-1)
PET-CT Scanners
0.094
PET-CT Scanners
Detector Dimension (mm)
# of PET Detectors
PET Detector Material
Spatial Resolution
2D/3D
Atten Corr
Detector Dimension (mm)
# of PET Detectors
PET Detector Material
Spatial Resolution
2D/3D
Atten Corr
GE Discovery STE
4.7 x 6.3 x 30
13,440
BGO
5.0
2D/3D
CT
GE Discovery ST
6.2 x 6.2 x 30
10,080
BGO
6.1
2D/3D
CT
Siemens Biograph LSO Siemens Hi-Rez LSO
6.5 x 6.5 x 25
4 x 4 x 20
9,216
23,336
LSO
LSO
6.3
4.6
3D
3D
CT
CT
Time-of-Flight PET
Philips Gemini
4 x 6 x 20
17,864
GSO
4.9
3D
CT&Cs-137
Speed of Light
Time (ns)
Distance (cm)
c = 3 x 1010 cm/s
0.1
0.5
1.0
5.0
3
15
30
150
8
Time-of-Flight PET
Time-of-Flight PET
x = c t/2
Speed of Light
Time (ns)
Distance (cm)
c=3x
1010
cm/s
Where x is the time-of-flight spatial
uncertainty and t is the timing resolution.
0.1
0.5
1.0
5.0
t (ns)
0.1
0.3
0.5
1.0
3
15
30
150
x (cm)
1.5
4.5
7.5
15.0
Time-of-Flight PET
Time-of-Flight PET
Assume t of 0.5 ns => x of 7.5 cm
9
Philips Gemini TF
PET scanner
Time-of-Flight PET
LYSO : 4 x 4 x 22 mm3
28,338 crystals, 420 PMTs
70-cm bore, 18-cm axial FOV
SNR Gain from Time-of-Flight PET
D/1.6 x
CT scanner
2 D/ 1.6 c t
Brilliance 16-slice
where D is the diameter of the object
D (cm)
20
30
40
SNR Gain
1.6
2.5
3.3
Performance measurements
Installation at U.Penn Nov ’05
Validation and research patient imaging
Nov ’05 - Apr ’06
50 patients
Beta testing and upgrade to production release software
May ’06 - Jun ’06
40 patients (to date)
Timing resolution = 600 ps
Courtesy of Joel Karp, PhD, Univ of Penn
Measurements: 35-cm lesion phantom
TOF
Intrinsic
non TOF
6-to-1 contrast
Energy resolution: 11.5% fwhm, Timing resolution: 585 ps
300s
NEMA NU-2
Spatial resolution: 4.8 mm at 1 cm, 5.2 mm at 10 cm
Sensitivity: 6.6 cps/kBq
180s
Scatter fraction (at 440 keV): 27% for 20-cm x 70-cm
Peak NEC: 125 kcps @ 0.42µCi/ml
Courtesy of Joel Karp, PhD, Univ of Penn
60s
Courtesy of
Joel Karp, PhD,
Univ of Penn
10
Gemini TF
Gemini TF - patient study
Rectal carcinoma, metastases in
mesentery and bilateral iliac chains
Heavy-weight patient study
13 mCi
2 hr post-inj
3 min/bed
Colon cancer
114 kg; BMI = 38.1
12 mCi; 2 hr post-inj
3min/bed
119 kg
BMI = 46.5
MIP
LDCT
non-TOF
non-TOF
TOF
TOF
Lesion contrast (SUV) improves with TOF reconstruction
Courtesy of Joel Karp, PhD, Univ of Penn
Design Criteria for MicroPET
• Image mice and rats
• Very high spatial resolution (~1-2 mm)
• High sensitivity (want to image with limited
amounts of radioactivity)
• Reasonable size to fit into a laboratory
• Access to animal about the device
Courtesy of Joel Karp, PhD, Univ of Penn
Spatial Resolution
• Positron range (related to positron
energy)
• Non-collinearity (uncertainty depends
on detector radius)
• Detector size
11
Positron Range
Positron Range
The distribution of
rms range is
exponential
and not well
characterized by
FWHM.
The rms range is proportional to positron energy.
From Cherry et al., Physics of Nuclear Medicine
Non-Colinearity
From Cherry et al., Physics of Nuclear Medicine
Limit on PET Spatial Resolution
(Regardless of detector size)
Rsys = SQRT[Rdet2 + Rrange2 + R1802]
The same level of non-colinearity
leads to a bigger uncertainty with a
larger detector diameter.
R180
0.002 x DR
Whole body scanner
(100 cm diameter)
18F
Small animal scanner
(20 cm diameter)
18F
82Rb
82Rb
~2.0 mm
~3.3 mm
~0.5 mm
~2.6 mm
12
Siemens MicroPET Focus 120 Scanner
•
•
•
•
•
•
•
Axial Field of View
Transaxial Field of View
Scintillating material
Number of Detectors
Detector Size
Spatial Resolution (CFOV)
Sensitivity
Siemens MicroPET Focus 120 Scanner
7.6 cm
10.0 cm
LSO
13,824
1.5x1.5x10 mm
1.7 mm
6.5%
650g Rat
[18F] Fluoride
4 bed positions
30 min each
Courtesy of Siemens Medical Systems
Siemens MicroPET Focus 120
Scanner
Imaging cholinergic receptors
in a mouse pancreas using microPET
Brain
Liver
Gated rat study
with FDG
Stomach
Pancreas
Bladder
Courtesy of Kathryn Morton MD
Wake Forest University
13
GE eXplore Vista Scanner
•Rodent system
•11.8 cm diameter
•4.6 cm axial FOV
•1.6 mm resolution
in CFOV
Philips Mosaic
•16,680 GSO crystals
•2.1 mm resolution
•11.8 cm axial FOV
•137Cs Transmission
Summary
• Modern scanners designed for oncologic imaging
• Practically all PET sales are PET/CT scanners
• New scintillation crystals combine excellent
detection efficiency with short decay times
• Shorter decay times leads to possibility of time-offlight PET.
• MicroPET scanners can provide very high spatial
resolution with high sensitivity in a small foot
print and easy access to the research animals.
14