Hi Filippo,
You could use "proportional normalisation" as the global normalisation procedure to adjust the voxel wise analysis for the variability in
injected dose of FDG and uptake times. This proportionally scales the voxel intensities to whole brain counts.
I don't think there is any way to account for the biological differences between the two groups due to the oncologic patients being able to have psychotropic medications and the neurologic patients having these withheld. Maybe you could look into the oncologic patients to see how many of these "control" subjects were taking such medications. It may not be a large proportion.
Good luck,
Kevin
PS/ I never thought of coffee as a psychotropic substance, but if it is then I am horribly addicted!
-----Original Message-----
From: SPM (Statistical Parametric Mapping) [mailto:[log in to unmask]] On Behalf Of Filippo A.
Sent: Saturday, 30 April 2016 05:33
To: [log in to unmask]
Subject: [SPM] PET acquisition difference and SPM study design
Hi all,
I’m here to request thoughts, comments ot tips from SPM experienced users/phisicians.
I’m new to SPM and I’m retrospectively studing neurologic (MCI or AD) patients comparing them, with a 2 sample T Test, to “healtly subjects” underwent to a 18F-FDG PET/CT.
The so-called “healtly-subjects” database came from oncologic patients undewent to PET/CT examination mainly for breast or lung tumors (TxNxM0).
Those “healtly-subjects” patients had a slightly different acquisition protocol from neurologic patients.
Brain scans for neurologic patients had a default acquisition time of 10 minutes, 20 minutes after about 185MBq FDG intravenous injection.
Brain scans for oncologics patients had a default acquisition time of 6 minutes, about 50 minutes after a 300-400MBq (patient weight and glycemia driven) intravenous injection.
Also, for neurologic patients psycotropics substances (like Tea or Cofee) withdrawal was mandatory.
1)Can those differences lead to inconsistent results ?
2)Did anyone already encountered this problem ?
3)Is there any “fix” (covariates or something) in the study design phase to minimize errors?
Thanks in advance for your help
Filippo
This message is intended for the addressee named and may contain confidential information. If you are not the intended recipient, please delete it and notify the sender.
Views expressed in this message are those of the individual sender, and are not necessarily the views of NSW Health or any of its entities.
|