NeuroRA

A Python Toolbox of Representational Analysis from Multimodal Neural Data

Overview

Representational Similarity Analysis (RSA) has become a popular and effective method to measure the representation of multivariable neural activity in different modes.

NeuroRA is an easy-to-use toolbox based on Python, which can do some works about RSA among nearly all kinds of neural data, including behavioral, EEG, MEG, fNIRS, sEEG, ECoG, fMRI and some other neuroelectrophysiological data. In addition, users can do Neural Pattern Similarity (NPS), Spatiotemporal Pattern Similarity (STPS), Inter-Subject Correlation (ISC) & Classification-based EEG Decoding on NeuroRA.

Paper

Lu, Z., & Ku, Y. (2020). NeuroRA: A Python toolbox of representational analysis from multi-modal neural data. Frontiers in Neuroinformatics. 14:563669. doi: 10.3389/fninf.2020.563669

Installation

pip install neurora

Documentation

You can read the Documentation here or download the Tutorial here.

Required Dependencies:

• Numpy: a fundamental package for scientific computing
• SciPy: a package that provides many user-friendly and efficient numerical routines
• Scikit-learn: a Python module for machine learning
• Matplotlib: a Python 2D plotting library
• NiBabel: a package prividing read +/- write access to some common medical and neuroimaging file formats
• Nilearn: a Python module for fast and easy statistical learning on NeuroImaging data
• MNE-Python: a Python software for exploring, visualizing, and analyzing human neurophysiological data

Features

• Calculate the Neural Pattern Similarity (NPS)

  for each subject / for each time-point / searchlight / for ROI

• Calculate the Spatiotemporal Neural Pattern Similarity (STPS)

  for each subject / searchlight / for ROI

• Calculate the Inter-Subject Correlation (ISC)

  for each time-point / searchlight / for ROI

• Calculate the Representational Dissimilarity Matrix (RDM)

  for each subject / for each channel / for each time-point / searchlight / for ROI / all in

• Calculate the Representational Similarity based on RDMs

  for each subject / for each channel / for each time-point / searchlight / for ROI / all in

• One-Step Realize Representational Similarity Analysis (RSA)

  for each subject / for each channel / for each time-point / searchlight / for ROI / all in

• Statistical Analysis

• Save the RSA results as a NIfTI file for fMRI

• Classification-based EEG Decoding

• Visualization results of representational analysis

Typical Visualization Demos

• Representational Dissimilarity Matrix

  

• Representational Similarities by time sequence

  

• RSA-result for fMRI

  

Script Demos to Know How to Use

There are two demos in Tutorial to let you know how to use NeuroRA to conduct representational analysis.

	Run the Demo	View the Demo
Demo 1
Demo 2
Demo 3

• Demo 1 for EEG/MEG, based on visual-92-categories-task MEG dataset, includes 8 sections.

  Section 1: Loading example data

  Section 2: Preprocessing

  Section 3: Calculating the neural pattern similarity

  Section 4: Calculating single RDM and Plotting

  Section 5: Calculating RDMs and Plotting

  Section 6: Calculating the Similarity between two RDMs

  Section 7: Calculating the Similarity and Plotting

  Section 8: Calculating the RDMs for each channels

• Demo 2 for fMRI, based on Haxby dataset, includes 8 sections.

  Section 1: Loading example data

  Section 2: Preprocessing

  Section 3: Calculating the neural pattern similarity (for ROI)

  Section 4: Calculating the neural pattern similarity (Searchlight)

  Section 5: Calculating the RDM for ROI and Plotting

  Section 6: Calculating the RDM by Searchlight and Plotting

  Section 7: Calculating the representational similarities between a coding model and neural activities

  Section 8: Saving the RSA result and Plotting

• Demo 3 for comparing classification-based decoding and RSA.

  Section 1: Downloading the data

  Section 2: Classification-based Decoding

  Section 3: Plotting the classification-based decoding results

  Section 4: RSA-based Decoding

  Section 5: Plotting the RSA-based decoding results

Users can see more details in Tutorial and Demo Codes.

About NeuroRA

Noteworthily, this toolbox is currently only a test version. If you have any question, find some bugs or have some useful suggestions while using, you can email me and I will be happy and thankful to know.

My email address: zitonglu1996@gmail.com

My personal homepage: https://zitonglu1996.github.io