Hi Expert
I have a question about resting-state analysis using meica.
I have 3T philips MRI in my university which can collect multi-echo EPI data.
But, it has limited working memory to storage data. So, it can save only 5 min data at one time. But, I need 15 min resting-state data to extract enough independent component by using meica analysis.
For example, meica can extract only 7 components from 5 min data, but it can extract 17 components from 15 min data.
Under these circumstances, I collected 5 min resting-sate data 3 times independently one by one.
Then, I did spatial normalization and de-meaned data by using meica preprocessing function (-pp_only). Then, I combined 3 independents data by using 3dTcat.
3dTcat -prefix sapmed-ALL.nii.gz meica.RS-fMRI-1-echo123/zcat_ffd.nii.gz meica.RS-fMRI-2-echo123/zcat_ffd.nii.gz meica.RS-fMRI-3-echo123/zcat_ffd.nii.gz
Then, I did meica analysis (tedana.py). It extracted independents components and denoised time course of resting state data (medn) which has combined 3 independent data.
Extracted components seems OK, but I am wondering whether this is right thing or not, because there is 2 artificial points connecting 3 independent data while they were demeaned and meica denoised drift to some extent.
Do you have any recommendation? Or, is there better way to analyze 3 independently collected resting-state with 3 TEs ?
Thank you.
Taka
