4/21/16 Concepts, Applications, and Requirements for Quantitative SPECT/CT Eric C. Frey, Ph.D. (efrey@jhmi.edu) Division of Medical Imaging Physics Russell H. Morgan Department of Radiology and Radiological Science Johns Hopkins University Conflict of Interest Disclosure Under 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. 1
4/21/16 Acknowledgements • People • Funding: NIH Grants – Bin He, Ph.D. (now at New – R01 EB 000288 York Hospital) – R01 CA 109234 – Yong Du, Ph.D. – R01 EB 000168 – Na Song, Ph.D. (Now at – U01 CA 140204 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 2
4/21/16 Single-Photon Emission Computed Tomography (SPECT) Gamma Camera γ-rays Imaging Agent Gamma Camera Gamma Camera Two-Camera SPECT/CT Systems X-ray Tube X-ray Detector 3
4/21/16 Gamma Cameras Matrix Formulation of Image Reconstruction 3D Activity Distribution Image Mean Projection Data Projection Measurement Noise Matrix (Quantum, Poisson Distributed) Reconstructed image (estimated activity distribution image) ‘Inverse’ of Projection Matrix Projection Matrix C is • Large (~4x10 12 elements) • Ill-Conditioned • Patient-dependent 4
4/21/16 Computed Tomography ( ) = ( ) δ y cos θ − x sin θ − t ( ) dxdy ∫∫ p t , θ a x , t y p ( t, θ ) a ( x,y ) θ x t This can be inverted analytically. The solution is known as Filtered Backprojection. Physical Image Degrading Factors • Attenuation } • Scatter • Collimator-Detector Response (CDR) Effects of – Geometric response high-energy emissions – Septal penetration and scatter responses • Partial Volume Effects • Statistical Noise 5
4/21/16 Ideal Projection from Point Source Ideal Collimator Source Attenuation in Patient Ideal Collimator Scattered Absorbed Source 6
4/21/16 Effects of Attenuation • Without attenuation compensation, sources at depth appear dimmer • Reduces quantitative accuracy Phantom FBP Reconstruction (no attenuation compensation) Object Scatter Ideal Collimator Unscattered Absorbed Source Multiply Scattered Scattered 7
4/21/16 Quantitative Effects of Scatter Scatter + Phantom Unscattered Unscattered Phantom Primary+Scatter Primary Reconstructed Intensity 3000.0 2500.0 2000.0 1500.0 1000.0 500.0 0.0 0 8 16 24 32 40 48 56 64 Pixel Number Collimator-Detector Response (CDR) Real Collimator Geometrically Septal Scatter Collimated Septal Penetration Source 8
4/21/16 Properties of the Full CDR 364 keV 30 cm I-131 Point Source MEGP Collimator HEGP Collimator Distance from 5 cm 10 cm 15 cm 20 cm Collimator Face T otal detected counts are a function of distance Effects of CDR on Spatial Frequencies • Analagous to spatially varying low-pass filter 1.2 Rectangular, ν m =0.5 1 LEHR 1.0 LEGP Butterworth, ν m =0.23, n=6 Relative Magnitude 0.8 0.8 Hann, ν m =0.5 0.6 GTF 0.6 0.4 0.4 0.2 0.2 0.0 0 0 0.5 1 1.5 2 0 0.1 0.2 0.3 0.4 0.5 Spatial Frquency (cm -1 ) Frequency (cycle/pixel) 9
4/21/16 Effect of CDR on SPECT Images Point Source FBP Reconstruction Phantom from Projections with LEHR Collimator Partial Volume Effects Phantom Reconstruction Spill Out 0.16 Image Intensity (Arbitrary Units) Spill In 0.14 Phantom 0.12 Reconstruction 0.1 0.08 0.06 0.04 0.02 0 0 50 100 150 200 250 Pixel 10
4/21/16 Statistical (Quantum) Noise Mean 2 kcounts 8 kcounts 32 kcounts 128 kcounts Described by Poisson Distribution Poisson Noise • Counts in pixels are independent 1.2 LEHR random variables 1.0 LEGP Noise power 0.8 • Noise has equal Spectrum: level GTF power at all 0.6 depends on counts in image frequencies 0.4 0.2 • Image has less 0.0 information at high 0 0.5 1 1.5 2 Spatial Frquency (cm -1 ) frequencies due to CDR 11
4/21/16 Effect of Poisson Noise on SPECT Images • Ramp filter used in FBP amplifies high frequencies • Combine with low-pass to reduce high this effect 0.5 Rectangular, ν m =0.5 Relative Magnitude 0.4 Ramp-Butterworth, ν m =0.23, n=6 0.3 Ramp-Hann, ν m =0.5 0.2 0.1 0 0 0.1 0.2 0.3 0.4 0.5 Frequency (cycle/pixel) 23 Effect of Poisson Noise FBP Reconstruction • Ramp filter amplifies high frequencies • Use low pass filter to reduce high frequency noise FBP w/ Noise Free FBP Ramp Ramp & Butterworth 24 12
4/21/16 Reconstruction-Based Compensation Project Initial Compare Computed Each Measured Estimate Computed & Projections Angle Projections Measured Model Cost Function New Update Estimate Estimate Applications of Quantitative SPEC/CT • Radiopharmaceutical Therapy Treatment Planning (absolute, lateral) • Diagnosis (relative, lateral) • Response to Therapy (relative, longitudinal) 13
4/21/16 Radiopharmaceutical Therapy (RPT) n Agents ( e.g., monoclonal antibodies, peptides, microspheres) that target tumors n Bound to radionuclides whose emissions can kill tumor cells n Crossfire effect n Bystander effect n Optimal dose is patient dependent n Treatment planning to determine administered activity Common Therapeutic Radionuclides for TRT β - Halflife γ Energy Radionuclide (hr) Energy (keV) (% yield) (MeV) 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), … 14
4/21/16 TRT Treatment Planning Flow Chart Administer Measure Calculate Planning Distribution Organ and Agent over Time Tumor Doses Administer Calculate Therapeutic Therapeutic Quantity Activity Cumulated Activity and Residence Time ( ) dt where τ = ∫ A = = A 0 τ A / A A t 0 t (MBq) A : Cumulated activity (MBq ⋅ sec) A 0 : Injected activity (MBq) Activity A(t) τ : Residence Time (sec) A Time t (sec) 15
4/21/16 SPECT/CT VOI Activity Estimation Measurement Estimate Sum SPECT Activity in SPECT Proj. Total VOI Reconstruction Liver & Convert to Activity Activity CT SPECT Residence Time Estimation SPECT 0.12 CT Proj. 0.1 0 hr 0.08 ( ) A t organ 0.06 A 0 0.04 SPECT 4 hr 0.02 Activity 0 Estimation 0 50 100 150 Time (hours) Curve Fitting 24 hr 0.12 0.1 72 hr 0.08 ( ) A t organ 0.06 A Residence 0 0.04 144 hr Time 0.02 0 0 50 100 150 Time (hours) 16
4/21/16 In-111 QSPECT Heart Lungs Liver Kidneys Spleen Reconstructed using 5% % Error in Residence Time Estimate OS-EM w/attenuation, scatter, CDR and (Estimated-True)/True *100% partial volume 3% compensation 1% 50 noise realizations Error bars show -1% standard deviations of activity estimates due to quantum noise -3% Precision better than -5% accuracy for most Heart Lungs Liver Kidneys Spleen Marrow QSPECT organs 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. Precision for Small Objects • 2.2 cm diameter tumors 0% Tumor 3 (2.2 cm, ratio 5.2) -2% -4% % Error in Activity Estimates -6% -8% -10% -12% Tumor 9 (2.2 cm, ratio 10.5) -14% T3 -16% T9 -18% -20% 0 5 10 15 20 25 30 35 40 45 50 # of Iterations (24 subsets/iteration) OS-EM w/attenuation, CDR and scatter compensation (no PVC) 17
4/21/16 Quantification of Very Small Objects • 0.9 cm diameter tumors -40% T2 Tumor 4 (0.9 cm, ratio 12) -45% -50% T4 % Error in Activity Estimates -55% -60% -65% -70% Tumor 2 (0.9 cm, ratio 11) -75% -80% -85% -90% 0 5 10 15 20 25 30 35 40 45 50 # 0f Iterations (24 subsets/iteration) OS-EM w/attenuation, CDR and scatter compensation (no PVC) I-131 Physical Phantom Philips Precedence SPECT/CT system with HEGP collimator Heart Large Small Myocardium Background Chamber Sphere Sphere Volume ( ml ) 17.5 5.7 59.7 115.3 9580 (r =1.61 cm) (r =1.11 cm) Activity( mCi ) 0.562 0.471 0.136 0.044 8.15 Activity concentration 9.38 4.08 7.77 7.72 0.851 ( mCi/μl ) § 128 projection views § Acquisition time: 40s / view 18
Recommend
More recommend
