What determines spatial resolution in SPECT and PET?
Spatial resolution in SPECT and PET is determined by different dominant physical factors.
In SPECT, spatial resolution is primarily limited by collimator geometry and source-to-detector distance. In PET, spatial resolution is limited by positron range, photon non-collinearity, and detector size.
SPECT spatial resolution is limited by collimator design, while PET spatial resolution is limited by positron range, photon non-collinearity, and detector size.
Although both are tomographic techniques, the physical mechanisms that define resolution differ fundamentally because SPECT uses mechanical collimation, while PET uses electronic coincidence detection.
Understanding the physics
In SPECT, radionuclides emit single gamma photons that must pass through a lead collimator before reaching the detector. The collimator restricts photon angles to provide spatial localisation. Smaller collimator holes and longer hole lengths improve resolution but reduce sensitivity. Resolution also worsens as the distance between the patient and collimator increases, because angular uncertainty translates into greater positional blur. As a result, SPECT resolution is dominated by mechanical design.
In PET, there is no mechanical collimator. Instead, coincidence detection defines a line of response between two detectors. This removes the resolution–sensitivity trade-off imposed by collimation. However, PET has its own intrinsic limits.
After decay, the emitted positron travels a short distance before annihilation. This positron range displaces the annihilation point from the original decay site. Higher-energy positrons travel further, worsening resolution. After annihilation, the two 511 keV photons are emitted nearly, but not perfectly, 180° apart. This slight deviation, known as non-collinearity, introduces additional spatial uncertainty that increases with scanner diameter.
Detector crystal size and reconstruction modelling further influence PET resolution, but even with ideal detectors, positron range and non-collinearity impose fundamental limits.
In summary:
SPECT resolution is dominated by collimator physics.
PET resolution is limited by emission physics and detector geometry.
Where this matters clinically
SPECT typically achieves spatial resolution of around 8–12 mm, while modern PET systems achieve approximately 4–6 mm. These differences influence small lesion detection and contribute to partial volume effects in both modalities.
Understanding the distinct resolution limits prevents over-interpretation of sub-centimetre lesions and clarifies why PET generally detects smaller lesions more reliably than SPECT.