/EEG-data-processing

A Python program to preprocess EEG(electroencephalogram) data (.edf format) which obtained from Biopac to get reletie power of each raw data

Primary LanguageJupyter Notebook

EEG-data-processing

Purpose

  • Program to preprocess EEG(electroencephalogram) data (.edf format) which obtained from Biopac to get reletie power of each raw data.
  • Reduce time effort for EEG data processing, mainly noise rejection and formatting
    • Previous workflow:

    Obtain data from acqKnowledge -> convert .acq EEG data to .edf format -> import raw data to eeglab (do preprocessing) -> downsampling rate -> IIR filtering, Notch noise filtering -> capture ECG pattern (noise) -> reject ECG noise from EEG data -> auto rejection with ASR algorithm -> export temp data -> process with software to get absolute power and relative band power for each data channel

    • Simpliszed workflow:

    Obtain data from acqKnowledge -> convert to .edf format -> run program -> get result

Requirements:

Use acqKnoledge software to obtain data from Biopac EEG 100 device.

  • data requirements

    .edf format EEG data data channle: not limited, same with MNE package

  • Environment

    • Python 3.7 +
    • Jupyterlab View/Edit notebooks of .ipynb format

  • Packages

    • Basic pacages

      • numpy==1.21.6
      • matplotlib==3.5.3
      • pandas==1.3.5

    • Algorithms

      • mne==1.3.1 exploring, visualizing, analyzing human neurophysiological data: MEG, EEG, sEEG, ECoG, NIRS ...
      • asrpy==0.0.2 automatic noise rejection

    • IO and data transform

      • eeglabio==0.0.2.post4 export data to .set format, which can be imported to eeglab
      • EDFlib-Python==1.0.8 export data to edf format
      • openpyxl==3.1.2 convert results into excel
      • XlsxWriter==3.1.8 write dataframe into excel

    • APP GUI and compiler

      • PySimpleGUI==4.60.5
      • psgcompiler==1.6.1

Directories and files

  • Folders

    • Raw data raw .edf data sample for test

    • notebook whole processing workflow and functions test for packages

      • EEG_Pre_Tets.ipynb test and debugging with MNE, ASR functions
      • EEG_single data processing_re.ipynb whole processing flow to handle a single edf data
      • EEG_multi data processing.ipynb modularized code to handle multi edf data

    • main_program_no_GUI .py files for multiple edf data processing, same functionality with ipynb files

    • GUI_Program add interface to make a program based on the previous code

    • main_GUI_ALL.zip download and unzip the compiled program, whole program running with Windows

Program guide

Run compiled program directly

Download main_GUI_ALL.zip, unzip and click main_GUI_ALL.exe to execute program Click browser to select target file folder, program will recognise edf files automatically

  • Select a single file to preview result, save result to local disk.
  • Process all edf data under current selected folder.

All processed results will appear under current selected folder with separate folder named after file name

  • .png Results of absolute and relative power of selected data
  • .edf edf data file after preprocessing, can be use as further aims
  • .set convert edf after preprocessing as a dataset of eeglab program, for further processing, you can import .set to eeglab directly and make use of that set file, for example, for further manual noise rejection
  • when process all data under folder, program will generate extra two xlxs files
    • absolute_power.xlsx absolute results for all data processed, separated with tab
    • relative_power.xlsx realative results for all data processed, separated with tab

Run whole project ⭐

  • OPT 1. Run non GUI python program. Clone folder named main_program_no_GUI, create a new python project in your IDE, change working directory to where your target data located, run main.py

  • OPT 2. Run GUI program. Clone folder named GUI_program, import to your IDE, run main.py

  • OPT 3. Run notebook. Download notebook in folder named notebook, open any ipynb support tools, for example Jupyterlab Import and run notebook to check

Compile program on other OS

You can download source code under > GUI_Program > final_file > main_GUI_ALL.py then compile with psgcompiler package :)

Program interface and functionality

main_interface.png

  • Browse: Open file explorer to select folder contains target edf files
  • Preview Selected : select a file and click to view results, show in right frame
  • Save Selected: Save the results of current file after preview results generated by program
  • Process All Files: Process all data under selected folder and generate results