Hi .
I have two questions related to ROI analysis:
Q1:
We would like to calculate mean percent change per entire ROI for different conditions.
Here is what we have done so far:
1. We have created the ROI mask for our area of interest (ROI_mask+orig).
2.We run the 3dDeconvolve and compute the percent change using the following formula:
3dcalc -fscale \
-a "decon_data[0]" -b "decon_data[2]" \
-c "decon_data[4]" -d "decon_data[6]" \
-e "decon_data[8]" -f "decon_data[10]" \
-g "decon_data[12]" -h "decon_data[14]" \
-i "decon_data[16]” \
-expr "i/(100*( a+0.5*b*num_scans_in_run1+ \
c+d*0.5*num_scans_in_run2 + \
e+f*0.5*num_scans_in_run3 + \
g+h*0.5*num_scans_in_run4 ) / num_runs)" \
-prefix PercentChange_cond1
3. We have run the command:
3dmaskave -mask ROI_mask+orig -mrange 1 1 -dindex 1 -drange 0 5 -sigma -dump -undump PercentChange_cond1+orig >> cond1.txt
My question refers to the significance of -drange a b ?
When we use -drange 0 1000 we get data about almost all voxels in our ROI. I wonder then what exactly means -drange 0 10 or -drange 0 5 and how can we interpret the results?;
Q2: What kind of other ROI analysis can be done except the following?
Mean Percent change per ROI.
Compute the average of voxels (ROI Average Plugin; 3dROIstats)
Extract timeseries information from an (ROI 3dmaskdump)
Creating a ROI based on a cluster in an activation map.
Thank you.