PET scanners -- SNR versus NECR....

  • Context: Graduate 
  • Thread starter Thread starter BobP
  • Start date Start date
  • Tags Tags
    pet Snr
Join the discussion
Ask a follow-up here, or get your own question answered by working scientists, mathematicians and engineers — people, not an autocomplete.
Real named experts · corrections over time · the nuance an AI answer skips
9 replies · 3K views
BobP
Messages
70
Reaction score
1
In PET, 2D PET scanners have a superior SNR to 3D PET scanners as the influence of scatter and random events is lowered by the septa between detector rings. However, the NECR (noise-equivalent count rate), which is linearly proportional to SNR, is higher in 3D PET scanners. These two facts seem to contradict each other and I am therefore slightly confused.

Please can someone clarify the situation for me
 
Physics news on Phys.org
BobP said:
...the NECR (noise-equivalent count rate), which is linearly proportional to SNR...
Shouldn't that be '...inversely proportional to SNR...' ?
 
Tom.G said:
Shouldn't that be '...inversely proportional to SNR...' ?
Not according to my lecturer...
 
BobP said:
However, the NECR (noise-equivalent count rate), which is linearly proportional to SNR, is higher in 3D PET scanners. These two facts seem to contradict each other and I am therefore slightly confused.

That sounds similar to Noise-Equivalent Temperature. Is NECR the number of signal counts needed to match the internal noise of the sensors?
 
Drakkith said:
That sounds similar to Noise-Equivalent Temperature. Is NECR the number of signal counts needed to match the internal noise of the sensors?
The NEC is the true count rate of a theoretical image without any scatter or random events which would give the same statistical quality image as a real image which includes scatter and randoms.
 
Okay. So it seems to me that the NECR for 3D PET is higher than that of 2D PET because the 3D PET real images are noisier and require more integration time to reach the same SNR as a 2D PET image.

In other words:

Real 2D PET
Signal: 100
Noise: 20
SNR: 5

Ideal 2D PET
Signal: 50
Noise: 10
SNR: 5
NECR: 50

Real 3D PET
Signal: 200
Noise: 40
SNR: 5

Ideal 3D PET
Signal: 100
Noise: 20
SNR: 5
NECR: 100I hope my terminology is okay, I'm used to talking about digital camera sensors and images, not PET sensors.
 
Drakkith said:
Okay. So it seems to me that the NECR for 3D PET is higher than that of 2D PET because the 3D PET real images are noisier and require more integration time to reach the same SNR as a 2D PET image.

In other words:

Real 2D PET
Signal: 100
Noise: 20
SNR: 5

Ideal 2D PET
Signal: 50
Noise: 10
SNR: 5
NECR: 50

Real 3D PET
Signal: 200
Noise: 40
SNR: 5

Ideal 3D PET
Signal: 100
Noise: 20
SNR: 5
NECR: 100I hope my terminology is okay, I'm used to talking about digital camera sensors and images, not PET sensors.
Well my lecturer said that SNR is basically the same as NECR. So it have a lower SNR and a higher NECR :)
 
BobP said:
Well my lecturer said that SNR is basically the same as NECR. So it have a lower SNR and a higher NECR :)

They certainly aren't the same, but you'll probably need to talk to your lecturer about this. Right now I'm mostly guessing at what all this means.
 
Ref post #2
Tom.G said:
Shouldn't that be '...inversely proportional to SNR...' ?
Ref post #3
BobP said:
Not according to my lecturer...
Ref post #7
BobP said:
Well my lecturer said that SNR is basically the same as NECR. So it have a lower SNR and a higher NECR :)

Posts #3 and #7 seem to conflict with each other. Please clarify.
 
Tom.G said:
Ref post #2

Posts #3 and #7 seem to conflict with each other. Please clarify.
Sorry. When I say "basically the same" I mean a high SNR = high NECR...
so they are linearly related...but I clarified this issue with my lecturer..

He said PET has a higher SNR and a higher NECR :)