Assuming that your input is timepoint2-timepoint1, your contrast will realize C = (INT2-INT1)-(CON2-CON1) as a final value.
On 9/6/2022 12:42 AM, rstein03@student.ubc.ca wrote:
External Email - Use Caution
Thank you for getting back to me. You are correct in your in that the contrast is C = (INT2-INT1)-(CON2-CON1). The measure of interest is SPC.
For the complete test we are trying to determine between group differences in cortical thickness, adjusting for covariates (age, sex, and MoCA). I was hoping to visualize this differencegraphically using the cache.th13.abs.y.ocn.dat file output.
This is our FSGD:
GroupDescriptorFile 1 Title Between-group Class int_male Class int_female Class con_male Class con_female Variables age moca Input 001.MR int_male 30 25 Input 002.MR int_female 40 22 Input 003.MR con_male 60 29 Input 004.MR con_female 60 21 … And this is the contrast matrix +0.5 +0.5 -0.5 -0.5 +0 +0 +0 +0 +0 +0 +0 +0
Thank you,
Ryan G. Stein
MSc Student, Rehabilitation Sciences
Aging, Mobility, and Cognitive Health Lab Djavad Mowafaghian Centre for Brain Health University of British Columbia | Vancouver Campus | Musqueam Traditional Territory
*From:* freesurfer-bounces@nmr.mgh.harvard.edu freesurfer-bounces@nmr.mgh.harvard.edu on behalf of Douglas N. Greve dgreve@mgh.harvard.edu *Sent:* August 28, 2022 12:45:22 PM *To:* freesurfer@nmr.mgh.harvard.edu *Subject:* Re: [Freesurfer] Group Analysis - glmfit results [*CAUTION:* Non-UBC Email]
Since this is longitudinal, I'm assuming that your contrast is actually something like contrast C = (INT2-INT1)-(CON2-CON1) where 1 and 2 the the time points. As you suggest, both (INT2-INT1) and (CON2-CON1) might both be less than 0, but if INT is less negative than CON, then C will be positive and everything is working as it should (ie, there is not a problem or discrepancy here). What is the (complete) test that you are trying to accomplish?
On 8/23/2022 3:10 PM, rstein03@student.ubc.ca wrote:
External Email - Use Caution
Hello Freesurfer Team,
I recently ran analysis to compare change in cortical thickness between an Intervention and Control group longitudinally. The contrast is set up as INT-CON, and therefore I would interpret a positive change as the intervention having a greater increase in cortical thickness or less atrophy than the control group.
The cluster that survived the Monte-Carlo simulation showed a positive difference in the LH superior parietal region when running both an ‘abs’ and ‘pos’ simulation (these showed the same cluster.summary outputs). There was not a significant cluster when running the ‘neg’ simulation. When I compared the mean change in thickness between groups (cache.th13.abs.y.ocn.dat file), it showed that the control group experienced less atrophy than the intervention despite having a positive cluster output.
One possible solution I considered is accounting for baseline differences between groups when looking at mean change. Does glmfit control for baseline cortical thickness in the cluster analysis that is not accounted for in the mean change values from the cache.th13.abs.y.ocn.dat file? Is there another possible explanation for this discrepancy?
Thank you in advance,
Ryan S
Ryan G. Stein
MSc Student, Rehabilitation Sciences
Aging, Mobility, and Cognitive Health Lab Djavad Mowafaghian Centre for Brain Health University of British Columbia | Vancouver Campus | Musqueam Traditional Territory
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu *MailScanner has detected a possible fraud attempt from "secure-web.cisco.com" claiming to be* https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer