07/11/2025 Meeting with Sonya Wang, MP Schallmo, Liz Kaplan, Sujin Park, and Sarah Chang

Action Plan

• Meet in 2-3 weeks to review changes to sliding window artifact rejection and ICA
  • Make changes to sliding window parameters
  • Test ICA implementation
  • Show data from revised changes to the pipeline

Meeting Notes:

Sliding window artifact rejection discussion

• Do we want to use this to remove blinks? No, just muscle artifact
• In MP’s lab, they use ICA to remove blinks, and the sliding window approach to remove vertical eye movement
• But for our purposes (removing muscle artifact), we want to change the parameters to specifically remove muscle and not eyeblinks
  • Epochs that are 5, 10, 1 second are likely too long for our purposes
  • Blinks take around 200 ms to occur
  • *** Increase amplitude for specifically muscle artifacts (100ms epoch size, 50 ms window size, step size 25 ms)

Premonitory urge discussion

• Some of the tics are blinks
• Double check that the blink tics are not totally discarded because of this
• All tics are timestamped, so we can pretty easily see this
• If we’re interested in the relative change during the premonitory urge, we should look at the time period before each premonitory urge
  • The kids’ subjective state and EEG background changes across 20 minutes of data acquisition (sometimes they’re more sleepy, sometimes more agitated)

ICA Discussion

• We have two data chunks (1) premonitory urge + non-overlapping tics (include first tic in a bout) (2) everything else
• Option 1: ICA => data chunk (1)
  • Note: ***You cannot apply IC weights from data chunk (1) to (2)
  • Apply IC weights to data chunk (1); then reinsert the data chunk (2)
• Option 2:
  • ???