This code repository maps out code used for the two papers below:
### (1) Developmental trajectories of EEG aperiodic and periodic components in children 2-44 months of age
Carol L. Wilkinson, Lisa Yankowitz,
Jerry Y. Chao, Rodrigo Gutiérrez, Jeff L. Rhoades, Shlomo Shinnar, Patrick L. Purdon & Charles A. Nelson
Paper can be found [here][1].
[1]: https://www.nature.com/articles/s41467-024-50204-4
### (2) Prediction of chronological age from resting-state EEG power in the first three years of life.
Winko W. An, Aprotim C. Bhowmik, Charles A. Nelson, Carol L. Wilkinson
Preprint can be found [here][2].
[2]: https://www.medrxiv.org/content/10.1101/2024.05.31.24308275v1
This code repository maps out the code used to (1) pre-process and completed multitaper power calculations from EEG data, (2) extract PSD from matlab BEAPP outputs and average across specified regions of interest, (3) run specparam/FOOOF and calculate aperiodic and periodic measures/features, and (4) complete analyses described in the paper. Below we describe the documentation and code stored in each folder.
#### `01_BEAPP_Input_parameters`:
* Matlab input parameters used to batch process EEGs using BEAPP/HAPPE 1.0.
* Word document with further description of parameters and electrode maps for 64 and 128 channel nets.
#### `02_Absolute_EEG_Power`:
Python code in this folder averages PSD data from BEAPP matlab outputs and averages across each segment/trial for each available electrode and frequency and creates a single CSV with data from a group of participants (`01_ConvertMatlabtoCSV_Example`).
`02_Averaging Power Data Across Electrodes for 64 and 128 channels – Example` Notebook is then used to average PSD data across regions of interest, remove data based on exclusion criteria or based on EEG quality metrics.
#### `03_Code_determining_EEG power_spectrum_measures`:
Python code in this folder were run in numbered order to:
`01_New_and_Old_FOOOF_Notebook_Trajectories`. Parametrize power spectra into aperiodic and periodic components using original and modified versions of SpecParam/FOOOF. Changes to the original SpecParam/FOOOF code are documented here.
`02_Merge_FOOOF_Spectra_Demo_info_Savgolfilter`: Merges spectral outputs from `01_New_and_Old_FOOOF_Notebook_Trajectories` notebook across studies and regions of interests and then performs smoothing using Savgol filter.
`03_Merge FOOOF_Aperiodic_metrics_Demo_info`: Merges spectral characteristic outputs (eg. aperiodic offset, aperiodic exponent) from `01_New_and_Old_FOOOF_Notebook_Trajectories` notebook across studies and regions of interests and merges in demographic data.
`04_Obtaining_FOOOF_power_bands`: From `02_Merge_FOOOF_Spectra_Demo_info_Savgolfilter` Notebook output, calculates power across various frequency bands.
`05_Identify_periodic_spectra_peaks`: From `Merge_FOOOF_Spectra_Demo_info_Savgolfilter` Notebook output, identified peaks and troughs within each periodic power spectra.
#### `04_Code_for_Developmental_Trajectory_analysis`:
Contains code used to create figures or run statistical analyses presented in the paper.
#### `05_feature_extraction_for_age_prediction`:
Contains code used extract EEG features for Age Prediction analysis.
#### `06_brain_age_prediction`:
Contains code used for predictive modeling and analyses in the Age Prediction Paper
