Print

Print
I’m not sure I know enough about your contrasts to understand what is going on.  Perhaps if you want to be more specific (off list if needed) I can be more helpful.  It is the case that head size may influence myelin map values through its effects on the residual transmit field.

Peace,

Matt.

From: FSL - FMRIB's Software Library <[log in to unmask]> on behalf of Lisa Kramarenko <[log in to unmask]>
Reply-To: FSL - FMRIB's Software Library <[log in to unmask]>
Date: Tuesday, August 22, 2017 at 11:56 AM
To: <[log in to unmask]>
Subject: Re: [FSL] Help with PALM results

as an additional info: I have run the same analysis just with two contrasts (without adjustment for age and gender) and the results are the same. It confuses me -- if in the first run there was such a big age effect, why do the results look the same when they weren't adjusted for age at all?

hope you have any ideas to help me further!

thanks!

On 22 August 2017 at 15:15, Lisa Kramarenko <[log in to unmask]> wrote:
Hi Matt,

I used the non bias corrected myelin maps as you said in the HCP mailing list that they are better for calculating group comparisons.

Best,
Lisa

On 22 August 2017 at 13:58, Matt Glasser <[log in to unmask]> wrote:
Which files did you use?

Peace,

Matt.

From: FSL - FMRIB's Software Library <[log in to unmask]> on behalf of Lisa Kramarenko <[log in to unmask]>
Reply-To: FSL - FMRIB's Software Library <[log in to unmask]>
Date: Tuesday, August 22, 2017 at 5:34 AM
To: <[log in to unmask]>
Subject: [FSL] Help with PALM results

Dear experts,

I have performed PALM analysis on CIFTI files (myelin maps from HCP). I used tail approximation and 500 permutations. I did it with the -logp option so that when I visualise the results I set the threshhold with the minimum at 1.3 (which corresponds to 0.05) and max to whatever max there is in the picture. Now my results seem a bit peculiar and I wanted to ask your advice.

I have six contrasts (g1>g2, g2>g1 and positive and negative effects of age and gender). In g1>g2 contrast literally almost whole brain is significant, whereas in the g2>g1 I cannot even threshhold anything because everything is zero. The positive effect of age, however, is also significant almost everywhere. 
Now I am a bit confused, a) if the positive effect of age is significant in almost all regions, should the g1>g2 comparison be not as strong (because it should be adjusted for the age effect?) Is it normal that g1>g2 still shows so much significance everywhere (same regions as the positive age effect)? and b) Even though it is line with my expectations I am a bit wary to see significant differences covering the whole brain. Is there maybe some error that might have led to that or something I have to double check? So far I checked my GLM and input files and cannot find anything weird/wrong there.

Sorry for the probably basic question and thanks a lot for the help!

Lisa