Granted Patent B2

US 12,450,795 · App. 18/079,370 · Granted Oct 21, 2025

Inventors: Timothy Deller (Brookfield, WI); Floribertus P. Heukensfeldt Jansen (Ballston Lake, NY); Matthew Gilbert Spangler-Bickell (Clovis, CA)

Assignee: GE PRECISION HEALTHCARE LLC

G06T11/008A61B6/037G06T11/005G06T2210/41G06T2211/424

Abstract

Claims (61)

1. A positron emission tomography (PET) system, comprising:

a detector ring assembly comprising a plurality of detectors, wherein each detector comprises one or more scintillation crystals and one or more photosensors; and

an image reconstruction computing device, comprising at least one processor electrically coupled with at least one memory device, and the at least one processor programmed to:

receive event data acquired by the PET system, wherein the event data are represented as a list of coincidence events, a coincidence event occurs when a positron emitted by radionuclide collides with an electron and a pair of photons are emitted due to collision and annihilation of the positron and the electron, and the pair of photons are detected by the plurality of detectors;

generate groups of coincidence events based on the event data, each group having a number of coincidence events in the event data and being representative of the event data;

perform a first level of image updates by iteratively updating a reconstructed image for a first number of times equal to a first number of image updates, wherein each image update is based on a first number of groups, the first number of groups including a first number of coincidence events;

perform a second level of image updates by iteratively updating the reconstructed image for a second number of times equal to a second number of image updates, wherein each image update is based on a second number of groups, the second number of groups including a second number of coincidence events, and the second number of groups greater than the first number of groups; and

output the reconstructed image.

2. The PET system of claim 1 , wherein the list of coincidence events are indexed with time indexes of the coincidence events, and the at least one processor is further programmed to:

perform motion correction based on the time indexes; and

output a motion-corrected image.

3. The PET system of claim 1 , wherein the at least one processor is further programmed to:

generate the groups of coincidence events by parsing the event data into the groups, wherein neighboring coincidence events are in different groups.

4. The PET system of claim 1 , wherein a difference in numbers of coincidence events per image update between consecutive levels is within a threshold level.

5. The PET system of claim 1 , wherein the at least one processor is further programmed to:

receive an update scheme defining a number of levels of image updates, a number of image updates at each level, and a number of coincidence events per image update at each level; and

for each level in the update scheme, the at least one processor is programmed to:

perform image updates by iteratively updating the reconstructed image for a number of times equal to a number of image updates at the level, wherein each image update is based on a number of coincidence events at the level.

6. The PET system of claim 5 , wherein the update scheme is user defined.

7. The PET system of claim 1 , wherein the at least one processor is programmed to:

perform the first level of image updates by iteratively updating the reconstructed image until a predetermined threshold is met.

8. The PET system of claim 1 , where the at least one processor is further programmed to:

generate the groups of coincidence events by:

estimating a percentage of randoms in the event data; and

determining the number of coincidence events in a group based on the percentage of randoms.

9. The PET system of claim 1 , wherein the at least one processor is programmed to:

randomize the event data by randomizing an order of the event data in time; and

perform image updates based on the randomized event data.

10. The PET system of claim 9 , wherein the at least one processor is further programmed to generate the groups of coincidence events by including an arbitrary number of coincidence events in the randomized event data into a group.

11. An image reconstruction computing device for a positron emission tomography (PET) system, comprising at least one processor electrically coupled with at least one memory device, and the at least one processor programmed to:

receive event data acquired by a PET system, wherein the PET system includes a detector ring assembly including a plurality of detectors, each detector includes one or more photosensors, the event data are represented as a list of coincidence events, wherein a coincidence event occurs when a positron emitted by radionuclide collides with an electron and a pair of photons are emitted due to collision and annihilation of the positron and the electron, and the pair of photons are detected by the plurality of detectors;

generate groups of coincidence events based on the event data, each group having a number of coincidence events in the event data and being representative of the event data;

perform a first level of image updates by iteratively updating a reconstructed image for a first number of times equal to a first number of image updates, wherein each image update is based on a first number of groups, the first number of groups including a first number of coincidence events;

perform a second level of image updates by iteratively updating the reconstructed image for a second number of times equal to a second number of image updates, wherein each image update is based on a second number of groups, the second number of groups including a second number of coincidence events, and the second number of groups greater than the first number of groups; and

output the reconstructed image.

12. The image reconstruction computing device of claim 11 , wherein the list of coincidence events are indexed with time indexes of the coincidence events, and the at least one processor is further programmed to:

perform motion correction based on the time indexes of the coincidence events; and

output a motion-corrected image.

13. The image reconstruction computing device of claim 11 , wherein the at least one processor is further programmed to:

generate the groups of coincidence events by parsing the event data into the groups, wherein neighboring coincidence events are in different groups.

14. The image reconstruction computing device of claim 11 , wherein a difference in numbers of coincidence events per image update between consecutive levels is within a threshold level.

15. The image reconstruction computing device of claim 11 , wherein the at least one processor is further programmed to:

receive an update scheme defining a number of levels of image updates, a number of image updates at each level, and a number of coincidence events in the event data per image update at each level; and

for each level in the update scheme, the at least one processor is programmed to:

perform image updates by iteratively updating the reconstructed image for a number of times equal to a number of image updates at the level, wherein each image update is based on a number of coincidence events at the level.

16. The image reconstruction computing device of claim 11 , wherein the at least one processor is programmed to:

perform the first level of image updates by iteratively updating the reconstructed image until a predetermined threshold is met.

17. The image reconstruction computing device of claim 11 , where the at least one processor is further programmed to:

generate the groups of coincidence events by:

estimating a percentage of randoms in the event data; and

determining the number of coincidence events in a group based on the percentage of randoms.

18. The image reconstruction computing device of claim 11 , wherein the at least one processor is programmed to:

randomize the event data by randomizing an order of the event data in time; and

perform image updates based on the randomized event data.

19. The image reconstruction computing device of claim 18 , wherein the at least one processor is further programmed to generate the groups of coincidence events by including an arbitrary number of coincidence events in the randomized event data into a group.

20. An image reconstruction computing device, comprising at least one processor electrically coupled with at least one memory device, and the at least one processor programmed to:

receive event data acquired by a medical imaging system, wherein the event data are represented as a list of coincidence events;

generate groups of coincidence events based on the event data, each group having a number of coincidence events in the event data and being representative of the event data;

perform a first level of image updates by iteratively updating a reconstructed image for a first number of times equal to a first number of image updates, wherein each image update is based on a first number of groups, the first number of groups including a first number of coincidence events;

perform a second level of image updates by iteratively updating the reconstructed image for a second number of times equal to a second number of image updates, wherein each image update is based on a second number of groups, the second number of groups including a second number of coincidence events, and the second number of groups greater than the first number of groups; and

output the reconstructed image.

Assignments (1)

Continuity (1)

References Cited (24)