4/21/16

1

Concepts, Applications, and Requirements for Quantitative

SPECT/CT Eric C. Frey, Ph.D. (efrey@jhmi.edu)

Division of Medical Imaging PhysicsRussell H. Morgan Department of Radiology and

Radiological ScienceJohns Hopkins University

Conflict of Interest DisclosureUnder a licensing agreement between the GE Healthcare and the Johns Hopkins University, Eric Frey is entitled to a share of royalty received by the University on sales of iterative reconstruction software used to obtain some results in this presentation.

Eric Frey is a co-founder of Radiopharmaceutical Imaging and Dosimetry, LLC. This company was founded to provide quantitative imaging and dosimetry service to developers of radiopharmaceutical therapy agents.

These interests have been disclosed and are being managed by the Johns Hopkins University in accordance with its conflict of interest policies.

4/21/16

2

Acknowledgements

• Funding: NIH Grants– R01 EB 000288– R01 CA 109234 – R01 EB 000168– U01 CA 140204

• People– Bin He, Ph.D. (now at New

York Hospital)– Yong Du, Ph.D.– Na Song, Ph.D. (Now at

Montefiore Medical Center)– Lishui Cheng– Xing Rong– Nadège Anizan, Ph.D.– George Sgouros, Ph.D.

Outline

• Introduction to SPECT• Applications of Quantitative SPECT• Requirements for Quantitative SPECT• Obstacles to Achieving Standardization

4/21/16

3

Single-Photon Emission Computed Tomography (SPECT)

γ-rays

Gamma Camera

Imaging Agent

Gam

ma

Cam

era

Gamma CameraTwo-Camera SPECT/CT Systems

X-rayTube

X-rayDetector

4/21/16

4

Gamma Cameras

Matrix Formulation ofImage Reconstruction

3D Activity Distribution Image

ProjectionMatrix

Mean Projection Data Measurement Noise(Quantum, Poisson Distributed)

Reconstructed image (estimated activity distribution image)‘Inverse’ of Projection Matrix

Projection Matrix C is • Large (~4x1012 elements)• Ill-Conditioned• Patient-dependent

4/21/16

5

Computed Tomographyp t,θ( ) = a x,t( )δ ycosθ − xsinθ − t( )dxdy∫∫

θ

y

x

p(t,θ)

t

a(x,y)

This can be inverted analytically. The solution is known as Filtered Backprojection.

Physical Image Degrading Factors

• Attenuation• Scatter• Collimator-Detector Response (CDR)

– Geometric response– Septal penetration and scatter responses

• Partial Volume Effects• Statistical Noise

}Effects of high-energyemissions

4/21/16

6

Ideal Projection from Point Source

Source

Ideal Collimator

Attenuation in Patient

Absorbed

Source

Ideal Collimator

Scattered

4/21/16

7

Effects of Attenuation

• Without attenuation compensation, sources at depth appear dimmer

• Reduces quantitative accuracy

Phantom FBP Reconstruction(no attenuation compensation)

Object Scatter

Scattered

Source

Ideal CollimatorUnscattered

MultiplyScattered

Absorbed

4/21/16

8

Quantitative Effects of Scatter

0.0500.0

1000.01500.02000.02500.03000.0

0 8 16 24 32 40 48 56 64

PhantomPrimary

Primary+Scatter

Pixel Number

Rec

onst

ruct

ed In

tens

ity

Phantom UnscatteredScatter +

Unscattered

Collimator-Detector Response (CDR)

Source

Real Collimator GeometricallyCollimated

Septal Penetration

Septal Scatter

4/21/16

9

Properties of the Full CDR

I-131 Point Source

MEGP Collimator

HEGP Collimator

Distance fromCollimator Face

5 cm 10 cm 15 cm 20 cm

30 cm

Total detected counts are a function of distance

364 keV

Effects of CDR on Spatial Frequencies

• Analagous to spatially varying low-pass filter

0.0

0.2

0.4

0.6

0.8

1.0

1.2

0 0.5 1 1.5 2

LEHRLEGP

GTF

Spatial Frquency (cm-1)

0

0.2

0.4

0.6

0.8

1

0 0.1 0.2 0.3 0.4 0.5

Rel

ativ

e M

agni

tude

Frequency (cycle/pixel)

Rectangular, νm

=0.5

Butterworth, ν

m=0.23, n=6

Hann, νm

=0.5

4/21/16

10

Effect of CDR on SPECT Images

Point SourcePhantom

FBP Reconstructionfrom Projections with

LEHR Collimator

0

0.02

0.04

0.06

0.08

0.1

0.12

0.14

0.16

0 50 100 150 200 250

PhantomReconstruction

Imag

e In

tens

ity (A

rbitr

ary

Uni

ts)

Pixel

Partial Volume EffectsPhantom Reconstruction

Spill Out

Spill In

4/21/16

11

Statistical (Quantum) Noise

Mean 2 kcounts 8 kcounts 32 kcounts 128 kcounts

Described by Poisson Distribution

Poisson Noise

• Counts in pixels are independent random variables

• Noise has equal power at all frequencies

• Image has less information at high frequencies due to CDR

0.0

0.2

0.4

0.6

0.8

1.0

1.2

0 0.5 1 1.5 2

LEHRLEGP

GTF

Spatial Frquency (cm-1)

Noise powerSpectrum: leveldepends on countsin image

4/21/16

12

23

Effect of Poisson Noise onSPECT Images

• Ramp filter used in FBP amplifies high frequencies

• Combine with low-pass to reduce high this effect

0

0.1

0.2

0.3

0.4

0.5

0 0.1 0.2 0.3 0.4 0.5

Rel

ativ

e M

agni

tude

Frequency (cycle/pixel)

Rectangular, νm

=0.5

Ramp-Butterworth, ν

m=0.23, n=6

Ramp-Hann, ν

m=0.5

24

Effect of Poisson NoiseFBP Reconstruction

• Ramp filter amplifies high frequencies• Use low pass filter to reduce high

frequency noiseNoise Free FBP Ramp

FBP w/Ramp & Butterworth

4/21/16

13

Reconstruction-Based Compensation

InitialEstimate

NewEstimate

ProjectEachAngle

ComputedProjections

MeasuredProjections

Model

UpdateEstimate

CostFunction

CompareComputed &Measured

Applications of Quantitative SPEC/CT

• Radiopharmaceutical Therapy Treatment Planning (absolute, lateral)

• Diagnosis (relative, lateral)• Response to Therapy (relative,

longitudinal)

4/21/16

14

Radiopharmaceutical Therapy (RPT)n Agents (e.g., monoclonal

antibodies, peptides, microspheres) that target tumors

n Bound to radionuclides whose emissions can kill tumor cellsn Crossfire effectn Bystander effect

n Optimal dose is patient dependent

n Treatment planning to determine administered activity

Common Therapeutic Radionuclides for TRT

RadionuclideHalflife

(hr)β-

Energy(MeV)

γ Energy (keV) (% yield)

I-131 192.5 0.6 0 364 (82), …Y-90 64 .0 2.28 none

Sm-153 46.3 0.81 103 (30), …Lu-177 161.5 0.50 208 (11), …Re-188 17.0 2.12 155 (15), …

4/21/16

15

TRT Treatment Planning Flow Chart

Administer Planning

Agent

Measure Distribution over Time

Calculate Organ and

Tumor Doses

Calculate Therapeutic

Activity

Administer Therapeutic

Quantity

Cumulated Activity and Residence Time

A : Cumulated activity (MBq ⋅sec)A0 : Injected activity (MBq)τ : Residence Time (sec)

A= A t( )dt

t∫ = A0τ

Act

ivity

A(t)

(MBq

)

A

Time t (sec)

where τ = A / A0

4/21/16

16

SPECT/CT VOI Activity Estimation

SPECT Reconstruction

& Convert to Activity

SumActivity in

VOISPECT

Proj.

CT

TotalLiverActivity

Measurement Estimate

SPECT Residence Time Estimation

ResidenceTime

CTSPECT

Proj.

Curve Fitting

SPECTActivity

Estimation

0 hr

4 hr

24 hr

72 hr

144 hr

0

0.02

0.04

0.06

0.08

0.1

0.12

0 50 100 150Time (hours)

( )0

A tA

organ

0

0.02

0.04

0.06

0.08

0.1

0.12

0 50 100 150Time (hours)

( )0

A tA

organ

4/21/16

17

In-111 QSPECT

-5%

-3%

-1%

1%

3%

5%

QSPECT

% E

rror i

n R

esid

ence

Tim

e Es

timat

e(E

stim

ated

-Tru

e)/T

rue

*100

%

Heart Lungs Liver Kidneys Spleen MarrowReconstructed usingOS-EM w/attenuation,scatter, CDR andpartial volumecompensation

50 noise realizations

Error bars showstandard deviations ofactivity estimates due to quantum noise

Precision better than accuracy for most organs

Heart Lungs Liver Kidneys Spleen Marrow

He B, Du Y, Song XY, Segars WP, Frey EC. A Monte Carlo and physical phantom evaluation of quantitative In-111SPECT. Phys Med Biol. 2005;50(17):4169-85.

-20%

-18%

-16%

-14%

-12%

-10%

-8%

-6%

-4%

-2%

0%

0 5 10 15 20 25 30 35 40 45 50# of Iterations (24 subsets/iteration)

% E

rror

in A

ctiv

ity E

stim

ates

Tumor 3 (2.2 cm, ratio 5.2)

Tumor 9 (2.2 cm, ratio 10.5)

• 2.2 cm diameter tumors

Precision for Small Objects

T9 T3

OS-EM w/attenuation, CDR and scatter compensation (no PVC)

4/21/16

18

-90%

-85%

-80%

-75%

-70%

-65%

-60%

-55%

-50%

-45%

-40%

0 5 10 15 20 25 30 35 40 45 50# 0f Iterations (24 subsets/iteration)

% E

rror

in A

ctiv

ity E

stim

ates

Tumor 4 (0.9 cm, ratio 12)

Tumor 2 (0.9 cm, ratio 11)

• 0.9 cm diameter tumors

Quantification of Very Small Objects

T2

T4

OS-EM w/attenuation, CDR and scatter compensation (no PVC)

§ 128 projection views§ Acquisition time: 40s / view

Heart Chamber Myocardium Large

SphereSmall

Sphere Background

Volume (ml) 59.7 115.3 17.5(r =1.61 cm)

5.7(r =1.11 cm)

9580

Activity(mCi) 0.562 0.471 0.136 0.044 8.15

Activity concentration(mCi/μl)

9.38 4.08 7.77 7.72 0.851

I-131 Physical PhantomPhilips Precedence SPECT/CT system with HEGP collimator

4/21/16

19

I-131 QSPECT

(%) Heart Large sphere(r = 1.61 cm

17.5 ml)

Small sphere(r = 1.11 cm

5.7 ml)AGS -15.21 -26.12 -32.72

ADS 4.75 -17.63 -25.77

ADS+Dwn+ -5.20 -21.10 -31.17

ADS+Dwn+PVC* -2.88 -15.49 -19.28

Percent errors of activity estimates for Anthropomorphic torso phantom

50 iterations 24 subsets/iteration

AGS ADS ADS + Dwn ADS+Dwn+PVE+DWN=model-based downscatter compensation*PVC=reconstruction-based PVC compensation

Y-90 QSPECT

38

• Physical phantom experiment– Elliptical phantom with 3 spheres– Philips Precedence SPECT/CT: HEGP– Acquisition time per view: 45s/view– Crystal thickness: 9.525 mm– 128 projection views over 360o

– Matrix size per view: 128*128– Pixel size: 4.664mm– VOIs defined from CT

38

4/21/16

20

Y-90 Physical Phantom Study

39

39Error = (EstimatedActivity – TrueActivity) / TrueActivity×100%

5.5 cm diametersphere

3.3 cm diameter sphere

1.5 cm diameter sphere

% Error -7.0% -9.7% -10.2%

Quantitative SPECT Imaging in Diagnosis and Monitoring of Parkinsonism

• Neurotransmission in dopaminergic systemØ Commercially available agent for SPECT (I-123

FP-CIT)

DAT D2RTatsch, Nucl. Med. Commun. (2001) 22, p819-827.

4/21/16

21

SPECT Imaging in ParkinsonismImaging assessment:Ø Visual inspectionØ Quantitative studies

üDisease degree, progression, etc.

Huang, et. al, Euro. J. Nucl. Med. (2004) 31, p155-161.

Normal Stage I PD Stage II PD

Stage III PD Stage IV PD Stage V PD

Quantitative brain SPECT Imaging

Huang, et. al, Euro. J. Nucl. Med. (2004) 31, p155-161.

Tc99m-TRODAT imaging of Parkinson’s disease in different HYS stages

4/21/16

22

Non-specific background uptake

Left putamenRight putamen

Left caudateRight caudate• GE Millennium VG/Hawkeye

(5/8” thick crystal)• LEHR Collimator• 128 views/360°, 128*128

projection w/ 0.24 cm pixels • CT attenuation maps• Manually defined VOIs using

registered MR Images• Activity concentrations:

• Bkg: 110 kBq/ml• Left Caudate: 212 kBq/ml• Left Putamen: 154 kBq/ml• Right Caudate: 1770 kBq/ml• Right Putamen: 222 kBq/ml

Accuracy of Activity Quantitation:I-123 Brain SPECT

RSD StriatalPhantom

Y. Du, B.M.W. Tsui, and E.C. Frey, "Model-based compensation for quantitative I-123 brain SPECT imaging," Phys Med Biol, 51(5): 1269-1282, 2006

Accuracy of Activity Quantitation:I-123 Brain SPECT

OS-EM w/AttenuationScatter &

CDRF Compensation Post-Reconstruction

pGTM PVC

4/21/16

23

Requirements for Quantitative SPECT/CT

• Quality Control/Calibration• Acquisition• Reconstruction/Processing

Quality Control & Calibration

• Activity meter calibration and QC• Routine Camera and CT QC• Registration of SPECT and CT• Calibration of QSPECT imaging

4/21/16

24

Calibration Factor Measurement• Planar calibration (sensitivity)

– Static image of standard source in air at known distance from camera

– Sensitivity = std. counts/(std. activity * acq. time)• Phantom-based Calibration

– Acquire SPECT study of object with known activity– Reconstruct and compute counts– Scale factor is true phantom activity/image counts– Should be consistent with planar calibration for “ideal”

reconstruction/compensation

Limitations of Planar CalibrationQuantitative Y-90 SPECT

• SPECT Calibration

Scanner Calibration Factor

GE Discovery 670

1.21-1.23

Siemens Symbia

1.15-1.18

Phantom DimensionsLarge Uniform Cylinder

20 cm diameter

Small Uniform Cylinder

4.6 cm diameter

Sphere in cold Elliptical Phantom

5.5 cmdiameter sphere in 32x20 phantom

• Planar CalibrationScanner Calibration

FactorGE Discovery 670

1.14

Siemens Symbia

1.08

4/21/16

25

Variations in Calibration Factor

1.8%

• Largest source of variation (77% of variance) was due to inter-source effects• Suggests that consistent preparation and measurement of source activity is key

Acquisition Parameters

• Collimator selection• Injected activity/acquisition time• Voxel size• Number of views• Energy windows

4/21/16

26

Reconstruction/CompensationFactor Large Object Small Object Commercially

Available

Attenuation Yes Yes Yes

Scatter Yes Yes Energy-based: yes

Model-based: limited

Geometric Response Compensation

No Yes Yes

Full CDR Compensation (High Energy)

Desirable forHE, ME

radionuclides

Desirable forHE, ME

radionuclides

No

Partial volume compensation

No Yes No

Noise Regularization

No Yes? Filtering

Obstacles to Standardization• Radioactivity measurement

– Variety of devices– Variety of radionuclides

• Compensation Methods– Many systems are SPECT-only– Variety of imaging hardware– Variety of image reconstruction and

compensation methods• Clinical Practice

– Variation in protocols– Habits

4/21/16

27

Summary

• Quantitative SPECT/CT is achievable now• There are a number of emerging clinical

applications• Limited commercial availability of state-of-

the-art reconstruction and compensation methods

• There is a need for standardization of protocols and methods