Print

Print
Hi Anderson,


Thanks for your help. As I am quite new to imaging, I hope you don't mind guiding me through these steps.

Let's say I have 1 PET data and 1 MRI of the same subject and the goal is to equalize their smoothing for further analyses using PALM.

I typed smoothest and these are the options that follow:

smoothest -d <number> -r <filename> -m <filename>

smoothest -z <filename> -m <filename>

Since the PET is a SUVR image (not 4D), I should use a z-stat image. What is that, and how may I get it to eventually fulfil my objective here?


On Thursday, August 10, 2017, 6:13:33 AM GMT+1, Anderson M. Winkler <[log in to unmask]> wrote:


Hi Shane,

Although it isn't strictly necessary, it is probably reasonable to have a similar degree of smoothing so that, for a given voxel, the contributions from the neighbouring ones are similar across modalities.

In FSL, the command "smoothest" can be used to yield an estimate of the smoothness of the residuals of (an initial fit) using the PET images, so that could offer some guidance on how much to smooth the MRI.

All the best,

Anderson


On 9 August 2017 at 12:20, Shane Schofield <[log in to unmask]> wrote:
Hi FSL Experts,

I would like to use PALM to look at inter-modality relationships (PET and MRI) and other patterns. Is it important to achieve the same level of smoothing before I do the analysis? I would like to compare things like Z scores (against another group) and voxel-to-voxel correlations. How can determine the formula to smooth my MRI to such a way that it is a smooth as PET?

Thank you.

Shane