Nuclear Medicine · Imaging physics

Gamma Camera, SPECT and PET Physics One-Liners

High-yield one-liners on gamma camera components, SPECT reconstruction, PET coincidence detection and key physics principles underlying nuclear medicine imaging.

Rapid-revision one-liners covering the physics of gamma camera detection, SPECT acquisition and reconstruction, PET coincidence detection and key quality-control principles.

Gamma camera detector
The gamma camera uses a thallium-activated sodium iodide [NaI(Tl)] scintillation crystal, typically 9.5 mm (3/8 inch) thick, which converts incident gamma photons into visible light detected by photomultiplier tubes.
Collimator function
The collimator is the primary determinant of spatial resolution and sensitivity in planar imaging; it absorbs obliquely incident photons, allowing only photons travelling perpendicular to the detector to form an image.
Low-energy high-resolution collimator
LEHR collimators have small holes and long septa, providing high spatial resolution at the expense of sensitivity; used for Tc-99m imaging at 140 keV.
High-energy collimator
High-energy collimators have thicker septa to prevent septal penetration by high-energy photons (e.g., I-131 at 364 keV); using an LEHR collimator for I-131 causes severe star artefact from septal penetration.
SPECT acquisition
SPECT acquires multiple planar projections over 180 or 360 degrees; typical protocols use 60-64 projections over 360 degrees with 20-30 seconds per projection.
Filtered back projection versus iterative reconstruction
Iterative reconstruction (e.g., OSEM) is now standard for SPECT, modelling photon attenuation, scatter and detector response; it produces fewer artefacts and better contrast recovery than filtered back projection.
SPECT/CT attenuation correction
The CT component provides an attenuation map (mu-map) for correction of photon attenuation, improving quantitative accuracy and reducing artefacts in deep structures such as the inferior wall on myocardial perfusion imaging.
PET coincidence detection
PET detects pairs of 511 keV annihilation photons arriving simultaneously (within 6-12 nanoseconds) at opposing detectors; the line connecting the two detection points is the line of response.
PET spatial resolution
Clinical PET spatial resolution is 4-6 mm FWHM, limited by positron range before annihilation, non-collinearity of annihilation photons (0.5 degrees deviation) and detector crystal size.
PET versus SPECT sensitivity
PET is 10-20 times more sensitive than SPECT because it uses electronic collimation (coincidence detection) rather than physical collimators, which reject over 99.9% of photons.
Time-of-flight PET
Time-of-flight (TOF) PET measures the small difference in arrival times of the two annihilation photons, localising the event along the line of response and improving signal-to-noise ratio, particularly in large patients.
PET attenuation correction
CT-based attenuation correction in PET/CT uses a bilinear transformation to convert CT Hounsfield units to 511 keV attenuation coefficients; PET/MRI uses segmentation or atlas-based methods.
Scatter correction
Compton scatter degrades contrast and quantitative accuracy; energy window-based methods (e.g., dual or triple energy window) and model-based iterative corrections are used in both SPECT and PET.
Dead time
At high count rates, detector electronics cannot process events fast enough (dead time), causing count loss and image degradation; PET scanners are more susceptible due to higher photon flux.
Uniformity quality control
Daily intrinsic uniformity flood with a point source or weekly extrinsic flood with a sheet source is required; non-uniformity greater than 3% coefficient of variation warrants investigation.
Centre of rotation
SPECT centre-of-rotation (COR) alignment must be verified regularly; COR misalignment causes ring artefacts and degrades resolution, particularly in small hot lesions.
Partial volume effect
Lesions smaller than twice the system spatial resolution (approximately 8-12 mm for PET) suffer underestimation of true activity due to partial volume averaging; correction algorithms are increasingly applied.
Random coincidences
Random coincidences occur when two unrelated photons are detected within the coincidence timing window; they increase with the square of activity and are estimated using a delayed coincidence window.

Caution

Specific resolution values, timing windows and QC thresholds vary by scanner manufacturer and model. Always reference institutional physics protocols and current IAEA or ACR quality-control guidelines.

Exam pearl

Remember that collimator choice is the dominant factor in SPECT resolution versus sensitivity trade-off, whereas PET resolution is fundamentally limited by positron range and non-collinearity.

Viva questions

Why is PET more sensitive than SPECT?
PET uses electronic collimation via coincidence detection, which does not reject the vast majority of photons as physical collimators do in SPECT. This gives PET approximately 10-20 times higher sensitivity.
What limits PET spatial resolution?
PET resolution is limited by positron range before annihilation, non-collinearity of the two 511 keV photons (approximately 0.5 degrees deviation from 180 degrees), and detector crystal element size. Clinical systems achieve 4-6 mm FWHM.
What happens if you image I-131 with a low-energy collimator?
Severe septal penetration occurs because I-131 emits 364 keV photons that penetrate the thin septa of a low-energy collimator, producing a star artefact and degrading image quality. A high-energy collimator with thicker septa is required.
How does time-of-flight PET improve image quality?
TOF PET measures the small arrival-time difference between the two annihilation photons, localising the event along the line of response rather than uniformly distributing it. This improves signal-to-noise ratio, particularly beneficial in large patients.
What is the partial volume effect and why does it matter clinically?
Lesions smaller than twice the system resolution (approximately 8-12 mm for PET) have their activity spread over a larger volume, causing underestimation of true SUV. This can lead to under-staging of small lesions or underestimation of treatment response.

Sources