search

Macquarie-MEG-Research / natural_conversations

Data analysis for the natural conversations study
0 stars 2 forks source link

ENH: Bidsification #4

Closed larsoner closed 11 months ago

larsoner commented 11 months ago

Progress toward BIDS:

  1. Set up BIDS dataset properties
  2. Add EEG electrode positions
  3. Align MEG and EEG and concatenate.
  4. Check drift rate and auditory jitter.

I tried using MNE-BIDS-Pipeline, but it has some drawbacks:

So maybe next steps instead would be:

  1. Fix timing of auditory localizer events in this script
  2. Add annotations to files of conversational segments in this script
  3. Rerun this script to have "finalized" data
  4. Analyze apart from MNE-BIDS-Pipeline for now

Thoughts @JD-Zhu ?

JD-Zhu commented 11 months ago

Thanks @larsoner !

The next steps sound good to me. I am currently using meegkit for TSPCA (can take lines 69-73 from simple_AEF_AEP.py). The improved timing for auditory localiser events can be obtained from utils.triggerCorrection() by passing in the raw MEG data, and the timestamps of conversational segments created by @yigu8115 can be read in using the align_and_segment_audios.load_processed_audio_data() function he provided.

JD-Zhu commented 11 months ago

Oops merging the PR seems to have closed the issue - let me know if I shouldn't have done that.

Also everyone should have direct write access as well :)