Dear Xin,
sorry for the naming confusion which is because of the serial runs of different modules in the longitudinal batch. Anyway, you can compare the baseline files of the longitudinal data with your cross-sectional data. They have passed the same preprocessing steps.
Regarding your study design, you have to keep in mind that it the limitation of longitudinal data for patient data only ignores the fact that you have some biological and technical noise in your control subjects. You have variations simply due to scanner instability and a bunch of biological effects (e.g. nutrition, hormones...) which are not considered by using longitudinal data only for one group. Furthermore, imagine the worst case that the coil of your scanner was changed between the acquisition of your baseline and your follow-up scans. Even, a different software release may cause differences. Thus, you can't differentiate anymore between these changes and the changes you are expect due to the disease.
Regards,
Christian
On Sun, 18 May 2014 16:12:07 -0400, Xin Di <[log in to unmask]> wrote:
>Dear Dr Gaser and SPMers,
>
>I am running a voxel based morphometry analysis on a group of patients with
>two repeated scans and a group of controls with only one scan. I have used
>the longitudinal pipeline of the VBM8 toolbox to analyzed the patients'
>data. The "wp1mrXXX_1st.nii" images were used for comparison between the
>patients' first and second scans. I am also interested in comparing
>differences between the control group and the first or second scans of the
>patient group. Because only one image was acquired for the control group, I
>can only used the VBM - Estimate and Write functionality to analyzed the
>control group data. My question is, which images are comparable with
>the "wp1mrXXX_1st.nii" images from the longitudinal pipeline. From my
>understanding, I should use the unmodulated normalized images, i.e. the
>'wrp1XXX.nii' images. Am I corrected? Thank you!
>
>Sincerely,
>XIn
>
>
>--
>Xin Di, PhD
>
>Research Assistant Professor
>Department of Biomedical Engineering
>New Jersey Institute of Technology
>University Height, Newark, NJ, 07102
>
|