External Email - Use Caution
Hello Freesurfer experts, Is there a way to constrain an analysis to a label or ROI in QDEC? I have a group analysis looking at the effect of inflammatory markers on thickness however, I’d like to have the analysis be constrained within an ROI rather than whole brain to increase the power. I specified a label in mir_glmfit but the analysis seems to just mask out the rest of the brain rather than constraining to the ROI. Thank you for all your help! Erin
No way in qdec. You can do it in mri_glmfit by specifying the --label option
On 8/18/2020 4:22 PM, Erin Moe wrote:
External Email - Use Caution
Hello Freesurfer experts, Is there a way to constrain an analysis to a label or ROI in QDEC? I have a group analysis looking at the effect of inflammatory markers on thickness however, I’d like to have the analysis be constrained within an ROI rather than whole brain to increase the power. I specified a label in mir_glmfit but the analysis seems to just mask out the rest of the brain rather than constraining to the ROI. Thank you for all your help! Erin
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
External Email - Use Caution
Hi Doug, Thanks for the prompt reply. I did use the label option in mri_glmfit, but it seemed to just show the same results as the whole brain analysis but for only in the region defined by the label. When I load the sig scalar into QDEC from the label analysis as well as sig scalar from the whole brain analysis, the results looks exactly the same within the region that I have labeled. Wouldn’t constraining the analysis to just the region defined by the label potentially increase the power and therefore the significance? When I compare the significant clusters from both analyses they show the same values. My apologies if this is not the case, I am a bit rusty with my stats. Thanks again for all the help! Erin
The uncorrected results will be exactly the same. The corrected results should be more significant
On 8/18/2020 7:44 PM, Erin Moe wrote:
External Email - Use CautionHi Doug, Thanks for the prompt reply. I did use the label option in mri_glmfit, but it seemed to just show the same results as the whole brain analysis but for only in the region defined by the label. When I load the sig scalar into QDEC from the label analysis as well as sig scalar from the whole brain analysis, the results looks exactly the same within the region that I have labeled. Wouldn’t constraining the analysis to just the region defined by the label potentially increase the power and therefore the significance? When I compare the significant clusters from both analyses they show the same values. My apologies if this is not the case, I am a bit rusty with my stats. Thanks again for all the help! Erin
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
freesurfer@nmr.mgh.harvard.edu
-
Douglas N. Greve -
Erin Moe