Thanks again. I hadn't considered using SIENA since we are estimating BOLD signal change in an fMRI paradigm rather than volumetric changes in anatomy. Is SIENA amenable to changes in data type?
By subtracting the two time points, do I understand you to indicate in isolation? Using FE model for t2-t1? One issue is that with attrition we have only about 2/3 the data points at t2 than we do at t1. The 1/3 of subjects who did not make it to t2 can still provide valuable information that I'd like to take into account. Similarly, I'd like to use the data for the subjects we have a third scan from at t3.
My point of comparison is to the lmer package for R, which works great for most data, but does not provide the more sophisticated estimates of variability that FSL does for imaging data.
Best,
David
|