/intersubj_rsa

An exploration of how to use inter-subject representational similarity analysis (is-RSA) to study individual differences in brain activity during naturalistic neuroimaging.

Primary LanguageJupyter NotebookMIT LicenseMIT

intersubj_rsa

This repository contains data and code associated with the following paper:

Idiosynchrony: From shared responses to individual differences during naturalistic neuroimaging
Emily S. Finn, Enrico Glerean, Arman Y. Khojandi, Dylan Nielson, Peter J. Molfese, Daniel A. Handwerker, Peter A. Bandettini
NeuroImage (2020). https://www.sciencedirect.com/science/article/pii/S1053811920303153

The code is contained in two Jupyter Notebooks:

  1. isRSA_simulations.ipynb reproduces the analyses shown in Fig. 2, which simulate different theoretical structures for brain-behavior representational similarity matrices.

  2. isRSA_HCP_demo.ipynb reproduces the analyses shown in Figs. 3 & 4, which show how inter-subject RSA can be empirically applied to movie-watching data from the Human Connectome Project. To run this notebook, you will need the individual-subject nodewise timecourses, which can be found in the folder all_shen268_roi_ts. You will also need the behavioral data from the HCP. The unrestricted behavioral data, which is freely available for download from the HCP, is included here for convenience (unres_behav_data.csv). The restricted behavioral data, which is necessary to take into account family structure when forming the two cohorts, requires signing a data-use agreement (DUA) and must be obtained through the HCP by individual investigators. See here for more information: https://www.humanconnectome.org/study/hcp-young-adult/document/restricted-data-usage. Once you have obtained the restricted data, you can save it as res_behav_data.csv to maintain compatibility with this notebook.