Skip to content

Latest commit

 

History

History
9 lines (7 loc) · 661 Bytes

README.md

File metadata and controls

9 lines (7 loc) · 661 Bytes

The "Podcast" ECoG dataset for modeling neural activity during natural story listening

To reproduce constructing this dataset, the following commands should be run in the following order:

  1. make bidsify: convert raw data from sourcedata/ into a BIDS dataset.
  2. make ecogqc: generate quality check reports in derivatives/ecogqc.
  3. make mark_bad_channels: based on the QC reports, this target will mark bad channels in _channels.tsv files.
  4. make ecogprep: run the raw data through the preprocessing pipeline.
  5. make generate_features: create various stimulus features based on stimuli/podcast.wav and stimuli/podcast_transcript.csv.