Emission Tomography (SPECT and PET) Instrumentation
Core concepts
SPECT uses a rotating gamma camera (or multi-head systems) with collimators; reconstruction is filtered back-projection (FBP) or iterative (OSEM). Attenuation correction may use a transmission source or CT. PET relies on annihilation producing two 511 keV photons emitted ~180° apart; coincidence detection within a timing window identifies true events. Detector materials (LSO, LYSO, BGO, GSO) differ in light yield and decay time. Time-of-flight (TOF) PET uses timing differences to localize events along the line of response. Artifacts include randoms, scatter, attenuation, and partial volume effects. Normalization, dead-time, and decay correction are standard corrections.
Traps
- Randoms increase with count rate (∝ activity²) while trues increase linearly - the random fraction worsens at high activity.
- Attenuation correction is not optional for quantitative PET; uncorrected images are nonuniform.
- Confusing coincidence timing window (nanoseconds) with energy window (511 keV ±).
- Partial volume effect underestimates activity in small lesions.
Rule of thumb
In PET, randoms ≈ 2τ·S₁·S₂ (τ = timing window); reducing the coincidence window reduces randoms but may lose trues.
Worked example: Doubling activity roughly doubles trues but quadruples randoms, so the noise-equivalent count rate degrades at very high activity.