Outline
PET Systems: Instrumentation and
Data Acquisition
• Basics of PET Scanner Design
• Review of Current PET Instrumentation
Frederic H. Fahey DSc
Children’s Hospital Boston
Harvard Medical School
frederic.fahey@childrens.harvard.edu
Positron Emission
Detector Ring
18F
+
511 keV
e511 keV
1
Detector Blocks (GE Advance NXi)
Large Crystal Designs
•Large NaI detectors, PMT array
•25x50 cm, 1” thick
•Hexagonal array
True, Scatter and Random
Coincidence Detections
New Detector Materials
SCINTILLATOR
Rel. Light Yield
True
Scatter
NaI(Tl)
100
Peak Wavelength (nm) 410
BGO
LSO
GSO
15-20
75
20-25
480
420
440
Decay Constant (ns)
230
300
12,42
30-60
Density (g/cc)
3.67
7.13
7.40
6.71
51
75
66
59
Index of Refraction
1.85
2.15
1.82
1.85
Hygroscopic ?
Yes
No
No
No
Effective Z
Random
(R = 2 τ N1 N2)
2
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
*
*
*
*
*
*
*
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
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
1
3D
septa
are
removed
*
*
*
*
*
*
*
16
16
2D
*
*
*
*
*
*
*
17
Span of 3 Michelogram
Acquisition Modes
*
*
*
*
*
*
*
18
* *
* * *
* * *
* * *
* * *
* * *
* * *
* * *
* * *
* * *
* * *
* * *
* * *
* * *
* * *
* * *
* * *
* *
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
2
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
3
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
4
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
5
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
6
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
7
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
8
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
9
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
10
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
11
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
12
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
13
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
14
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
15
Segment 2
16
Segment 1
Segment 3
3
PET Attenuation Correction
3D PET
P2 = e-µ(L-x)
•
•
•
•
•
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
X
P1 = e-µx
PET Attenuation Correction Methods
• Calculated
– No noise but possibly inaccurate
• Measured
– Accurate but noisy
• Segmented, Measured
– Less noise => less time
• Singles
• CT
L
PTOT = P1 x P2
= e-µL
Measured Attenuation Correction
100
20
Blank
C
Transmission
Emission
Corrected Emission = (Blank/Transmission) * Emission
=
(100/20) * C
4
Segmented, Measured
Attenuation Correction
•Noise added from measured attenuation
correction
•Rel err in unif phantom (10 min EM)
•9% with calc atten
•16% with 10 min TR
•18% with 5 min TR
•Segmentation classifies by tissue type
•Smoothes lung areas
•Substantial reduction in noise added
Measured
Singles-Based Attenuation
Correction
•Single-photon source (Cs-137)
•Source shielded from
“near”detectors
•Energy resolution (NaI) allows
separation of 2 peaks
•High count rate reduces noise
•Susceptible to scatter
Segmented
PET-CT Attenuation Correction
PET-CT Attenuation Correction
5
PET Instrumentation
•
•
•
•
High-Resolution
Medium-Resolution
PET-CT
Special Purpose
GE Advance NXi
PET Instrumentation
• High-Resolution
– Siemens HR+
– GE Advance
– Philips Allegro
• Medium-Resolution
– Siemens ACCEL
– Philips CPET+
PET Instrumentation
Detector Dimension (mm)
# of Detectors
Detector Material
Spatial Resolution (mm)
Sensitiv ity (kcps/uCi/mL)
Siemens HR+
4.1 x 4.4 x 30
18,432
BGO
4.6
200/900
GE Advance
Philips Allegro
3.9 x 8.2 x 30
4 x 6 x 20
12,096
17,864
BGO
GSO
4.8
4.8
200/1060
/800
Detector Dimension (mm)
# of Detectors
Detector Material
Spatial Resolution (mm)
Sensitivity (kcps/uCi/mL)
Siemens Accel
6.8 x 6.8 x 20
9,216
LSO
6.2
180/780
Philips CPET+
500x300x25
6
NaI
5
/450
6
GE Discovery ST
PET-CT Scanners
•
•
•
•
Siemens Biograph
GE Discovery ST
GE Discovery STE
Philips Gemini
CT
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
PET
Special PET Devices
Philips Gemini
4 x 6 x 20
17,864
GSO
4.9
3D
CT&Cs-137
• Brain
• Breast
• Small Animal
7
Positron Emission Mammography
(PEM)
CTI ECAT HRRT
• Max Planck Institute in
Cologne
• Dual LSO Phoswich for
DOI Determination
• Octagonal Design
• 936 blocks with 128 2.1x2.1
dual detectors each
• 120,000 crystals
• Reconstructed resolution of
less than 2.5 mm
Courtesy of CTI
Concorde MicroPET Scanner
650g Rat
[18F] Fluoride
4 bed positions
30 min each
Courtesy of Wake Forest University and PEM Technologies
Summary
• Scanners specifically optimized for clinical
(oncologic) imaging
• Sensitivity and count rate capability
• New detector materials (faster!)
• PET-CT
• Special devices including small animal
imaging
Courtesy of Concorde MicroSystems
8