Dear Rick,
I used the command that:
3dcalc -a Run1_smooth+orig -b Run1_mean+orig -c Run_all_mean+orig -expr 'c * min(200, a/b*100)' -prefix Run1_3dcalc
As I understand, it means that I scaled to a mean of 100.
In my experiment, there were three runs; they were combined to one using 3dTcat; and then I performed the regression analysis.
In fact, I don’t exactly about running waver and using the ‘-peak 1’ option.
I could not explain well about the C.
In my experiment, there were three conditions (A: intrinsic motivation, B: incentive motivation, C: non-motivation).
The tasks of three conditions were almost same and the only difference was details of phrases (e.g., A: interesting task, B: rewarding task, C: unsatisfactory task).
Because my main interest was to recognize the difference between A and C, I only compare A-C and B-C in the subtraction method (t-tests).
That’s why I described C as a baseline.
But, I think it’s not the baseline in terms of analyzing the brain data, right?
The fixation of this experiment was just presenting ‘+’.
So, I assumed that it was totally different from three conditions.
And also, I used 3dmerge with 5 FWHM.
As I understand, my problem was that I used very small constant volumes of each region for the ROI analyses.
In my analyses, I think I need to do creating ROI datasets from activation maps (of the “Region of Interest (ROI) Drawing” in the educational material in the AFNI website).
But, I am not sure how to draw ROI regions (e.g., ACC, anterior precuneus) and calculate mean beta values of each condition even after recognizing activated regions using 3dclust.
Could you let me know about the ROI method in this situation?
Thanks.
Woogul Lee