you cannot do this in QDEC. In fact, it is a tricky analysis to do properly. For each group you will need an offset regressor and an age regressor, but only the patient group will have a burden regressor. This cannot be done with an FSGD file either. You will have to create your own design matrix, then pass it to mri_glmfit with the --X flag. You can get part of the way with an FSGD file (or QDEC) if you remove the Burden factor. This will create a design matrix to which you can add a single column for Burden which takes the Burden value for patients but is 0 for controls.

doug

On 12/05/2013 12:40 PM, Cat Chong wrote:

Dear FreeSurfer Experts, We are trying to run a between group analysis (1 patient group 1 control group) within qdec. We are interested in investigating age-related thickness differences between groups. Our fixed factors are ‘group’ and our continuous covariate is ‘age’. We would like to control for other possibly confounding variables which should be declared as “nuisance factors, is that correct? Here is our problem. We like to put down “disease burden” (years subject has had the disease) as our nuisance factor, since it is likely possible that this has an effect on cortical thinning in our disease group. Yet, when we run the model, qdec crashes. Is this because the subjects in the control group have a “0” as disease burden? Could you advise us on another way of controlling for disease burden? Cheers, Catherine

---

Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer