NMT Boards – PET/CT & Auxiliary Equipment

Energy and number of photons produced by positron annihilation

Two 511 keV photons

Angle between the two annihilation photons

~180° (180° ± 0.25°)

What PET uses instead of a lead collimator

Electronic collimation (coincidence detection)

Line connecting two detectors that register a coincidence

Line of response (LOR)

Typical PET coincidence timing window

~4–12 nanoseconds

Positron range blur of F-18 (RMS)

~0.2 mm

Positron range blur of Rb-82 (RMS)

~2.6 mm

Why long positron range (Rb-82) worsens PET resolution

PET images annihilation site, not decay site

Coincidence event where both unscattered photons from one annihilation are detected

True coincidence

Coincidence event where one or both photons Compton-scatter before detection

Scatter coincidence

Coincidence event from two unrelated annihilations paired by chance

Random (accidental) coincidence

Coincidence event where 3 or more detectors trigger in one window

Multiple coincidence (usually rejected)

Formula for random coincidence rate

R = 2τ × N1 × N2

Randoms scale with injected activity how

With the square of activity

PET prompts formula

Prompts = trues + randoms + scatters

PET delayed window used for randoms estimation

60–100 ns

Corrections applied to PET data

Randoms, attenuation, scatter, dead time, decay

BGO crystal decay time

~300 ns

LSO crystal decay time

~40 ns

NaI(Tl) crystal decay time

230 ns

Light output of BGO relative to NaI

15%

Light output of LSO relative to NaI

75%

PET crystal with the highest stopping power

BGO

PET crystals that enable time-of-flight

LSO and LYSO

Dopant used in LSO, LYSO and GSO crystals

Cerium

Properties of an ideal PET crystal

Dense, high Z, high light output, short decay time

Solid-state PET photodetector replacing PMTs

SiPM (silicon photomultiplier)

What time-of-flight PET measures

Arrival-time difference between the two photons

Modern TOF timing resolution

~450–500 picoseconds

Patients who benefit most from TOF PET

Larger patients

Does TOF replace attenuation correction

No, still needs CT or MR attenuation map

Sensitivity gain of 3-D (septa out) vs 2-D PET

4–6× higher

Scatter fraction in 2-D vs 3-D PET

~10–15% vs ~30–40%

Three fundamental PET resolution limits

Positron range, crystal size, non-collinearity

Non-collinearity blur scales with

Ring diameter

Typical PET vs SPECT spatial resolution

~4 mm vs ~12 mm

PET sensitivity compared to a gamma camera

50–100× more sensitive

PET bore diameter

56–70 cm

PET axial FOV per bed position

15–18 cm

Bed positions for a whole-body PET

5–8

Half-life of C-11

20.4 min

Half-life of N-13

9.96 min

Half-life of O-15

123 s

Half-life of F-18

110 min

Half-life of Ga-68

68 min

Half-life of Rb-82

76 s

Half-life of Ge-68

271 days

Daily PET QC scan with no patient in the field

Blank scan

Source used for the PET blank scan

Ge-68 rod or cylinder (sealed)

Duration of the daily PET blank scan

5–10 minutes

Blank scan sinogram shows a thin dark diagonal line

Single faulty crystal

Blank scan sinogram shows a broad diagonal band

Failed detector block or PMT

PET scan that corrects per-LOR sensitivity variation

Normalization scan

Frequency and duration of PET normalization

Quarterly, ≥6 hours

Appearance of poorly normalized PET data

Ring or streak artifacts

Typical PET energy window for LSO

~425–650 keV

PET calibration that converts counts to Bq/mL for SUV

Well-counter (absolute activity) cross-calibration

PET cross-calibration must use

The same dose calibrator used for patient doses

Relationship of scanner calibration error to SUV error

About 1:1

Annual PET performance testing standard

NEMA NU-2

Frequency of PET coincidence timing check

Weekly

PET artifact causing a cold band at the lung base/diaphragm

Respiratory misregistration

PET artifact with a bright rim where the patient exceeds the CT FOV

Truncation artifact

PET artifact from metal or dense contrast

False hot spots (overcorrection)

Images to review when a PET/CT artifact is suspected

Non-attenuation-corrected (NAC) images

PET acquisition mode recording every event with a timestamp

List mode

Gating that extracts the signal from list-mode data without hardware

Data-driven gating

8-gate vs 16-gate cardiac PET LVEF compared to echo

8 underestimates; 16 matches

Respiratory gating more robust to irregular breathing

Amplitude-based gating

4D PET means

Respiratory-gated PET

Bed motion giving uniform axial sensitivity

Continuous bed motion (CBM)

SUV formula

Tissue conc ÷ (injected dose ÷ body weight)

Body weight unit used in SUV

Grams

SUV of a tracer distributed uniformly with no excretion

1

Normal liver SUV (internal control)

~1.7–2.7

Effect of FDG dose infiltration on SUV

Falsely low SUV

SUV error from using kg instead of g

1,000× error

SUV example: 70 kg, 10 mCi at scan time, lesion 0.0007 mCi/mL

SUV ≈ 4.9

FDG uptake time before imaging

~60 min

FDG mechanism of localization

Glucose analog, trapped as FDG-6-phosphate

Adult FDG dose

10–20 mCi (370–740 MBq)

Pediatric FDG dose

0.10–0.14 mCi/kg, min 0.7 mCi

FDG patient fasting requirement

4–6 hours, water only

Common US blood glucose limit for FDG PET

200 mg/dL reschedule

Effect of insulin given too close to FDG injection

Diffuse skeletal muscle uptake

CT parameter controlling beam quality, penetration and contrast

kVp

CT parameter controlling beam quantity, noise and dose only

mA (mAs)

CT noise relationship to mAs

Noise ∝ 1 ÷ √mAs

mAs relationship to dose

Linear

CT pitch formula

Table feed per rotation ÷ beam collimation width

Pitch with 12 mm collimation and 18 mm table feed

1.5

Effect of pitch >1

Faster, lower dose, worse z-axis resolution

Effect of thinner CT slices

More noise, less partial volume

Whole-body effective dose of AC-only vs diagnostic CT

~3–6 mSv vs ~11–20 mSv

CT filter that evens the beam across the patient

Bow-tie filter

CT collimators that set slice thickness

Pre-patient collimators

CT noise from too few x-rays

Quantum mottle

HU formula

1,000 × (µtissue − µwater) ÷ µwater

HU of water

0

HU of air

−1,000