Skip to content

README.md

Latest commit

 

History

History
45 lines (34 loc) · 2.08 KB

README.md

File metadata and controls

45 lines (34 loc) · 2.08 KB
Raw

SleepProfiler

This repository contains the functions necessary for processing the eeg data derived from SleepProfiler headbands. To perform analyses on your data derived from SleepProfiler device, you should run the SM_main_eeg_preprocess.m file.

#Prior to running this function, you need:

  1. Have eeglab installed. (and have its path added to your MATLAB)
  2. Have statistics toolbox installed
  3. Have the "Dependencies" folder added to your MATLAB path WITH SUBFOLDERS
  4. Have the "csc-eeg-tools-develop" folder added to your MATLAB path WITH SUBFOLDERS
  5. Have SPM12 installed. (and have its path added to your MATLAB) The inputs for this function are:
  6. subjects id
  7. source path: the path to your data (subject folder)
  8. dep_path: the path you have saved the dependency functions (on default, it should be in ...\DREEM3\Dependencies)
  9. epoch duration
  10. sesssion number : night number

%%%%%%%%%%%%%%%EXAMPLE INPUTS: %%%%%%%%%%%%%%%clc %%%%%%%%%%%%%%%clear all %%%%%%%%%%%%%%%sub_id = 'COV263'; %%%%%%%%%%%%%%%source_path = '/home/soodeh/Desktop/SleepProfiler/data/RawData'; %%%%%%%%%%%%%%%dep_path = '/home/soodeh/Desktop/SleepProfiler/SleepProfiler_Scripts/Dependencies'; %%%%%%%%%%%%%%%ses_num = 'ses-03'; %%%%%%%%%%%%%%%epoch_duration =30;

What does SM_main_eeg_preprocess.m do?

  1. Converts the .edf file to .set file. which is a structure that can be read by eeglab toolbox.
  2. Adds the lables to the sleep stages.
  3. Saves the .set file with sleep stages
  4. Pops up a visual control for the first time(you can check the overal quality of the whole data-REM and NREM- here)
  5. Creates a .set structure for N2N3 (which represent the NREM sleep)
  6. Detects and delete the bad signals using "wispic" method as implemented in eeglab
  7. visually check the data after artifact rejection only for NREM sleep
  8. check if you have even number of clips in deleting the bad signals

Note: You do not need to clip the signal at step 4, the code takes care of the signal processing later. but you HAVE TO at least clip a small portion at step7 (this is to make sure that you checked the data by eye at some point)