Paul Tesla
Fri. 9 Nov.07,
15:31
|
|
I know this is a basic question but how do you measure SNR?
This page:
http://dnl.ucsf.edu/users/dweber/dweber_docs/mri_quality.html
tells me to measure an as small as possible ROI for the signal mean and as large as possible ROI for the standard deviation.
But I have a book called "MRI from picture to proton": This says that the standard deviation of the noise (Sn) can be measured using the mean or standard deviation (sd) of a background region. Then you can use the relationship:
Mean = 1.25 Sn
sd = 0.66 Sn
to find Sn. But where do the values 1.25 and 0.66 come from?
It also says that the measured mean of the signal ROI is related to the MR signal instensity (S) using:
Mean = sqrt(S^2 + Sn^2)
But why do they add up quadratically?
Thankyou for your help. Please leave a message or email me,
Paul
-------------
mri.tesla@gmail.com
|
Forum Overview
Category:
