PET Instrumentation and Physics
What this domain tests
NMTCB PET questions in this domain probe how annihilation photons are detected, how acquisition mode and detector material change image quality, and how corrections and reconstruction shape the final SUV image. Expect calculations and conceptual comparisons rather than wiring diagrams.
Core rules
- A positron loses kinetic energy over a short range, then annihilates with an electron, emitting two 511 keV photons about 180° apart. Non-collinearity (~0.5°) and positron range blur resolution; Ga-68 and Rb-82 have longer ranges than F-18.
- A coincidence is accepted when both photons hit within the timing window (often 4–12 ns) and fall in the energy window (roughly 350–650 keV). Narrower energy windows reject scatter but reduce sensitivity.
- Randoms rise with activity squared and with timing-window width; delayed-window correction estimates them. Scatter rises in 3D mode, which removes septa and raises sensitivity but also randoms and scatter.
- LSO/LYSO offer fast, bright signals; BGO has high stopping power but lower light output. Smaller crystals improve spatial resolution but reduce sensitivity.
- TOF uses arrival-time differences to localize the annihilation along the line of response, improving SNR, especially in large patients. OSEM is standard iterative reconstruction; CT-based attenuation correction is needed for quantitative SUV.
Worked example
At 370 MBq, randoms are 5% of trues. Activity doubles to 740 MBq. Trues double, but randoms quadruple: new randoms fraction = (4×0.05T)/(2T + 4×0.05T) = 0.20T/2.20T ≈ 9.1%. This is why high-activity or 3D studies need robust randoms correction.
Traps
Do not confuse positron range with non-collinearity; both degrade resolution but originate differently. Do not assume 3D always improves image quality-it increases scatter and randoms. Remember that CT attenuation correction can introduce artifacts from metal or misregistration.