For the power analysis, you need 3 or 4 things:
1. effect size (mean/std)
2. false positive rate (usually .05)
3. Number of subjects (or design matrix and contrast)
4. true positive rate (usually .8)
Given any 3, you can compute the 4th. For the FS output, #1 should be gamma.mgh/rstd.mgh (rstd.mgh is in the glm folder, and gamma is in the contrast folder). We do not really have any code to do the power analysis itself.
On 12/11/17 1:42 PM, John Anderson wrote:
Dear Freesurfer exerts, I read sentences similar to this one in multiple publications (e.g. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3500581/) "Power analyses were carried out using vertex-wise variance estimates from the coregistered cortical thickness maps, systematically varying processing parameters. "
Please how can I use freesurfer to do power analysis. Cortical thickness (vertex-based analysis) is very well documented in Freesurfer wiki. How can I use the output of this analysis (i.e. gamma.mgh and the gammavar.mgh ) to measure the power.
I highly appreciate any advice. John
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer