An axial section shows two rows of detector rings with the patient between; annihilation photons leave at random axial angles. In 2D the septa between rings stop oblique and scattered photons; in 3D coincidences between all rings are accepted.

Step 1/2

2D (septa)

Lead or tungsten septa sit between the rings. Photons travelling obliquely hit a septum and are absorbed; only pairs reaching the same or an adjacent ring are recorded. Most photons scattered in the patient also stop in the septa: scatter and randoms are low, but because most photons are lost, sensitivity is low.

2D acquisition (with septa)patientCoincidences accepted: 12septa (lead/tungsten)axial direction →
2D and 3D compared2D3DBetween ringssame/adjacentallSensitivitylow~4–8×Scatter fraction~15–20%~30–40%RandomslowhighReconstructionsimplecomplextrue LORscattered (wrong LOR)Schematic; axial direction stretched. Values from the book.
Continuous stream · no seeking

Key point: removing the septa lets cross-ring coincidences count and raises sensitivity ~4–8×; the price is more scatter and randoms and a more complex reconstruction.

Source: book (2D: same/adjacent ring, scatter ~15–20%; 3D: sensitivity ~4–8×, scatter ~30–40%, randoms↑). The drawing is schematic: the axial direction is stretched, so the ratio of counts in the drawing does not show the real gain. Photon directions isotropic; single scatter of ~30% of photons in the body is illustrative.

How to use: a single click on the drawing pauses and resumes; ← → switch between 2D and 3D.

Did this animation help you understand the topic?

2D and 3D PET acquisition · Open in Nuclear Medicine Explained ↗