PET Scan (Positron Emission Tomography)

Introduction

PET: functional imaging technique detecting metabolic activity using positron-emitting radiotracers. Distinction: CT/MRI show anatomy, PET shows function (metabolism, receptor binding, blood flow). Primary tracer: FDG (fluorodeoxyglucose), glucose analog taken up by metabolically active cells. Applications: oncology (80%), cardiology, neurology. Market: ~5,000 PET/CT scanners worldwide. Impact: fundamentally changed cancer staging and treatment monitoring.

"PET reveals the invisible chemistry of life and disease. While CT shows the anatomy of a tumor, PET shows whether it is alive and growing, dormant, or dying. This functional information transforms clinical decision-making." -- Nuclear medicine physician

Physics of Positron Emission

Positron Decay

Mechanism: proton-rich nucleus converts proton to neutron, emitting positron (β⁺) and neutrino. Positron: antimatter electron (same mass, opposite charge). Range: travels 1-3 mm in tissue before annihilation (limits spatial resolution). Energy: kinetic energy varies (0 to maximum, tracer-dependent).

Annihilation

Event: positron meets electron → both destroyed. Product: two 511 keV gamma photons emitted in opposite directions (180° apart). Detection: coincidence detection of both photons identifies annihilation location. Line of response (LOR): line connecting two detectors that fire simultaneously.

Common Positron Emitters

Radiotracers

FDG (18F-Fluorodeoxyglucose)

Structure: glucose analog with F-18 replacing hydroxyl group. Uptake: transported into cells by glucose transporters (GLUT). Phosphorylation: hexokinase traps FDG-6-phosphate intracellularly. No further metabolism: accumulates proportional to glucose utilization. Cancer: high glucose metabolism → high FDG uptake (Warburg effect).

PSMA Tracers

Target: prostate-specific membrane antigen (overexpressed in prostate cancer). Tracers: Ga-68 PSMA-11, F-18 DCFPyL (Pylarify), F-18 PSMA-1007. Application: prostate cancer staging, biochemical recurrence. Advantage: far more sensitive than conventional imaging for prostate cancer. Impact: changing management of prostate cancer worldwide.

Somatostatin Receptor Tracers

Target: somatostatin receptors (overexpressed in neuroendocrine tumors). Tracer: Ga-68 DOTATATE (Netspot). Application: neuroendocrine tumor staging and treatment planning. Theranostic: paired with Lu-177 DOTATATE for peptide receptor radionuclide therapy (PRRT).

Amyloid and Tau Tracers

Amyloid: F-18 florbetapir (Amyvid), flutemetamol (Vizamyl), florbetaben (Neuraceq). Tau: F-18 flortaucipir (Tauvid). Application: Alzheimer's disease diagnosis and staging. Impact: confirm amyloid pathology in vivo (previously only at autopsy). Emerging: tau PET correlates better with clinical symptoms.

Tracer Production

Cyclotron: accelerates protons to bombard target (nuclear reaction produces radioisotope). Radiochemistry: attach isotope to biological molecule (automated synthesis modules). Quality control: purity, sterility, endotoxin testing. Distribution: F-18 tracers shipped within ~2 hours (regional distribution). On-site: Ga-68 generator (no cyclotron needed).

Detection System

Scintillation Detectors

Crystal: BGO (bismuth germanate), LSO (lutetium oxyorthosilicate), LYSO. Function: convert 511 keV gamma photon to visible light pulse. Photosensor: photomultiplier tube (PMT) or silicon photomultiplier (SiPM). Energy resolution: 10-15% (distinguishes 511 keV from scattered photons). Timing resolution: 300-600 ps (determines time-of-flight capability).

Coincidence Detection

Principle: two detectors fire within coincidence time window (4-10 ns). Valid coincidence: both photons from same annihilation. Random coincidence: two unrelated photons detected simultaneously (noise). Scattered coincidence: one or both photons scattered before detection. Correction: random subtraction, scatter correction algorithms.

Detector Ring Geometry

Full ring: detectors surround patient (360° coverage). Axial extent: 15-25 cm (conventional), 70-194 cm (total-body PET). Number of crystals: 20,000-50,000+ per scanner. 3D acquisition: all lines of response (no septa), maximum sensitivity. Data rate: millions of coincidences per second.

Time-of-Flight (TOF)

Principle: measure arrival time difference between two photons. Information: localize annihilation along LOR (not just on line). Timing resolution: 200-400 ps (localizes to ~3-6 cm). Advantage: improved signal-to-noise ratio (√gain factor). Impact: better image quality, especially for large patients.

Image Reconstruction

Sinogram Data

Raw data: coincidence events organized by angle and position. Sinogram: 2D representation of projection data. Correction: attenuation, scatter, random, normalization, dead time. Attenuation correction: critical (511 keV photons significantly attenuated in body).

Attenuation Correction

Problem: photons absorbed in body (deeper structures appear less active). Solution: CT-based attenuation map (PET/CT). Process: CT Hounsfield units converted to attenuation coefficients at 511 keV. Impact: essential for quantitative PET (without correction, images are non-quantitative).

Reconstruction Algorithms

FBP: fast but noisy. OSEM (Ordered Subsets Expectation Maximization): iterative, standard clinical method. PSF modeling: includes system resolution in reconstruction (sharper images). Bayesian: regularized reconstruction (controls noise). AI: deep learning approaches emerging.

Resolution and Sensitivity

Spatial resolution: 3-5 mm (limited by positron range and detector size). Sensitivity: ~1-5% (fraction of annihilation events detected). Total-body PET: 40x higher sensitivity (detects nearly all emitted photons). Voxel size: 2-4 mm (clinical), 1 mm (research).

PET/CT Hybrid Imaging

Concept

Combined scanner: PET + CT in single gantry. Acquisition: CT first (seconds), then PET (15-30 minutes). Fusion: automatically registered (same bed position). CT provides: anatomic reference + attenuation correction. PET provides: functional/metabolic information. Standard: virtually all clinical PET performed as PET/CT.

Advantages

Localization: FDG uptake precisely localized to anatomy. Attenuation correction: CT-based (faster, more accurate than transmission scan). Efficiency: single examination (no separate CT appointment). Incidental findings: CT may reveal additional pathology. Standard of care: replaced standalone PET.

Protocol

Patient preparation: fasting 4-6 hours (lower blood glucose). FDG injection: ~370-555 MBq IV. Uptake time: 60 minutes (quiet rest, minimize muscle uptake). Scan: vertex to mid-thigh typical (20-30 minutes). CT: low-dose for attenuation correction, or diagnostic quality if clinically needed.

CT Dose Considerations

Low-dose CT: ~3-5 mSv (attenuation correction only). Diagnostic CT: ~10-15 mSv (full diagnostic quality). PET radiation: ~7 mSv (from injected FDG). Total: ~10-25 mSv per PET/CT examination. Optimization: use lowest CT dose adequate for clinical question.

PET/MRI

Technical Challenge

Problem: conventional PMTs don't work in magnetic field. Solution: silicon photomultipliers (SiPMs) are MRI-compatible. Attenuation correction: no CT available (MRI-based methods required). MRI-based AC: Dixon-based tissue segmentation (less accurate than CT). Status: limited installations (~200 worldwide).

Advantages over PET/CT

Soft tissue contrast: MRI superior to CT (brain, liver, pelvis). Radiation reduction: no CT radiation dose. Simultaneous acquisition: PET and MRI acquired together (motion correction). Multi-parametric: PET + diffusion + perfusion + spectroscopy. Application: brain, pediatrics, liver, pelvis.

Challenges

Cost: $4-7M (vs. $2-3M for PET/CT). Throughput: slower (MRI takes longer than CT). Attenuation correction: less accurate than CT-based. Lung/bone: MRI inferior to CT for these tissues. Adoption: limited to academic centers.

Oncology Applications

Cancer Staging

FDG PET/CT: standard for staging many cancers. Lymphoma: initial staging and response assessment (Deauville criteria). Lung cancer: staging, treatment planning (mediastinal involvement). Head and neck: primary tumor, nodal metastases, unknown primary. Melanoma: detect distant metastases. Colorectal: recurrence detection.

Response Assessment

PERCIST criteria: PET Response Criteria in Solid Tumors. SUV measurement: quantify metabolic response. Complete metabolic response: no residual FDG uptake. Partial response: >30% decrease in SULpeak. Progressive disease: >30% increase or new lesions. Timing: interim PET (mid-treatment) guides therapy changes.

Radiation Therapy Planning

GTV delineation: PET defines metabolically active tumor volume. Advantage: more accurate than CT alone (identifies viable tumor within mass). Application: lung cancer, head and neck, lymphoma. Integration: PET/CT simulation for treatment planning. Impact: reduces treatment volume, spares normal tissue.

FDG-Negative Cancers

Low uptake: renal cell carcinoma, prostate cancer, hepatocellular carcinoma. Alternative tracers: PSMA (prostate), F-18 choline (HCC), Ga-68 DOTATATE (NET). Limitation: FDG not universal cancer tracer. Solution: expanding tracer menu for specific tumor types.

Cardiac PET

Myocardial Perfusion

Tracers: Rb-82 (generator-based), N-13 ammonia (cyclotron). Protocol: rest and stress (pharmacologic with regadenoson/adenosine). Quantification: absolute myocardial blood flow (mL/min/g). Advantage over SPECT: higher resolution, attenuation correction, quantitative flow. Indication: intermediate-high risk chest pain, known CAD assessment.

Myocardial Viability

FDG: glucose metabolism indicates viable (hibernating) myocardium. Protocol: glucose loading + FDG injection + cardiac PET. Pattern: FDG uptake without perfusion = hibernating (viable, will recover with revascularization). Impact: guides revascularization decisions (bypass vs. medical therapy). Evidence: strong evidence for improved outcomes when viable myocardium revascularized.

Flow Reserve

Coronary flow reserve: stress/rest flow ratio. Normal: CFR > 2.0. Abnormal: CFR < 2.0 (suggests significant coronary disease). Microvascular disease: diffusely reduced CFR without focal stenosis. Advantage: quantitative assessment not possible with other modalities.

Neurological PET

Dementia Evaluation

FDG: glucose hypometabolism pattern indicates dementia type. Alzheimer's: temporoparietal hypometabolism. Frontotemporal: frontal/temporal hypometabolism. Lewy body: occipital hypometabolism. Amyloid PET: confirms amyloid pathology (positive in Alzheimer's). Tau PET: correlates with disease severity.

Epilepsy

Interictal FDG: seizure focus shows hypometabolism. Application: presurgical localization (where to operate). Sensitivity: ~80-90% for temporal lobe epilepsy. Ictal SPECT: complementary (shows hypermetabolism during seizure). Impact: improves surgical planning, increases seizure-free outcomes.

Brain Tumors

FDG: limited by high normal brain glucose metabolism. Amino acid tracers: F-18 FET, C-11 methionine (better tumor-to-background contrast). Application: differentiate recurrence from radiation necrosis. Grading: high-grade tumors show higher amino acid uptake. Treatment planning: define biological tumor volume.

Quantification and SUV

Standardized Uptake Value (SUV)

SUV = tissue activity (Bq/mL) / (injected dose (Bq) / body weight (g))SUVmax: maximum voxel value in region of interestSUVmean: average value in ROISUVpeak: average in 1 cm³ sphere centered on hottest voxelNormal range: liver ~2-3, brain ~5-8, blood pool ~1.5-2.0

Factors Affecting SUV

Blood glucose: hyperglycemia reduces FDG uptake (compete for GLUT). Uptake time: longer = higher SUV (more accumulation). Body composition: lean body mass correction (SUL) preferred. Scanner calibration: cross-calibration with dose calibrator essential. Reconstruction: parameters affect SUV (standardization needed).

Kinetic Modeling

Dynamic PET: time-activity curves from serial frames. Compartment models: quantify tracer kinetics (influx rate Ki, distribution volume). Patlak analysis: graphical method for irreversible tracers (FDG). Logan plot: reversible tracer analysis. Application: research, quantitative pharmacokinetics.

Emerging Technologies

Total-Body PET

Concept: detector covers entire body simultaneously (194 cm axial). Scanner: uEXPLORER (UC Davis/United Imaging). Sensitivity: 40x higher than conventional PET. Impact: ultra-low dose imaging, rapid scans (30 seconds), delayed imaging, total-body kinetics. Application: pediatric (dose reduction), pharmacokinetics, whole-body dynamic imaging.

Theranostics

Concept: same molecule for diagnosis (PET) and therapy (therapeutic isotope). Example: Ga-68 DOTATATE (imaging) → Lu-177 DOTATATE (therapy) for NET. PSMA: Ga-68/F-18 PSMA (imaging) → Lu-177 PSMA (therapy) for prostate cancer. Impact: personalized treatment selection based on imaging.

AI in PET

Reconstruction: deep learning reduces noise (lower dose imaging). Attenuation correction: AI-generated attenuation maps from PET data alone. Segmentation: automatic tumor delineation. Outcome prediction: radiomic features predict treatment response. Workflow: automated SUV measurements, report generation.

Novel Tracers

Immuno-PET: radiolabeled antibodies image immune checkpoint expression. Fibroblast activation protein (FAP): pan-cancer tracer (higher sensitivity than FDG for some cancers). Hypoxia tracers: F-18 FMISO identifies radiation-resistant hypoxic tumors. Cell tracking: radiolabeled cells track immunotherapy response.

References

• Cherry, S. R., Sorenson, J. A., and Phelps, M. E. "Physics in Nuclear Medicine." Elsevier, 4th ed., 2012.
• Warburg, O. "On the Origin of Cancer Cells." Science, vol. 123, no. 3191, 1956, pp. 309-314.
• Boellaard, R., Delgado-Bolton, R., Oyen, W. J. G., et al. "FDG PET/CT: EANM Procedure Guidelines." European Journal of Nuclear Medicine and Molecular Imaging, vol. 42, no. 2, 2015, pp. 328-354.
• Badawi, R. D., Shi, H., Hu, P., et al. "First Human Imaging Studies with the EXPLORER Total-Body PET Scanner." Journal of Nuclear Medicine, vol. 60, no. 3, 2019, pp. 299-303.
• Hofman, M. S., Emmett, L., Sandhu, S., et al. "Lu-PSMA-617 versus Cabazitaxel in Progressive Metastatic Castration-Resistant Prostate Cancer (TheraP)." Lancet, vol. 397, no. 10276, 2021, pp. 797-804.