Hi Alfonso,

I would like to extract the timeseries of a specific ROI to correlate with behavioural score. After going through the forum, my understanding is that I should find them in the ROI_Subject*_Condition*.mat files.
My files contain data for 740 timepoints, which correspond to the number of scans per subject. However, given that these data (e.g.ROI_Subject*_Condition*.mat) are obtained after 'preprocessing', and thus the estimation of invalid scans, how can I make sure I am not using the time series of the 'bad' scans?

For instance, subject 1 had 740 scans before preprocessing. After preprocessing, 250 scans were considered as invalid (because of movement, and thus contaminated with noise). The ROI_Subject1_Condition1 is still giving me the time series of the 740 scans, although 250 were considered as invalid after preprocessing.

Should I not only focus on the time series of the 'valid' scans? Is there a way to control for that?

Thank you very much.