Let’s Get Started

In these introductory chapter we will give you a rapid overview of examples of some of the capabilities of SignalFlow and how they can be used to process your EEG data.

SignalFlow in a Nutshell

At it’s essense, SignalFlow is a MATLAB toolbox and provides a unified wrapper around opensource EEG toolkits (e.g., EEGLAB, FieldTrip, MNE, etc.) to process data and generate uniform outputs (tables and figures). Signalflow can handle EEG data from human recordings but also from other signal sources such as ?sec-mouse-eeg.

This flexibility should allow you to chose the best toolkit for a specific task, while still providing uniform outputs that can be analyzed similarly.

An Brief Example Workflow

Let’s give a prototype example of a SignalFlow workflow that might be difficult in any single EEG toolkit.

  1. MATLAB: Use the EEGLAB import functions to properly import an Magstim/EGI .mff file.

  2. MATLAB: Use a visual pipeline to preprocess several subjects using EEGLAB’s preprocessing functions including using Artifact Subspace Reconstruction (ASR) to remove artifacts.

  3. MATLAB: Use an optimized parallel ICA function to run ICA on all subjects in parallel using a custom ICA binary that is optimized for your operating system.

  4. MATLAB: Following preprocessing, use EEGLAB to save the files as SET files in a BIDS compatible format and save preprocessing details as a CSV file.

  5. Use the artifact-free SET to perform analyzses across toolkits but create R-statistics (CSV) compatible tables.

    1. Use MATLAB to generate band power tables
    2. Use Python to generate connectivity tables
    3. Use R to import the tables and run statistics and perform visualization.

We will go through variations of this workflow in the following sections to give a broad overview of the capabilities of SignalFlow.