Changes in the brain after psilocybin therapy for depression

This repo contains code for the statistical analysis and figure generation associated with the results described in: https://www.nature.com/articles/s41591-022-01744-z

BLUF: The evidence from two independent studies suggests that psilocybin therapy can alleviate depression via an increase in the brain's functional network integration.

Aims and usage

1. To extract measures of brain network modularity (Q) from preprocessed resting-state fMRI.
2. Compare the Q of patients with depression before and after psilocybin therapy.
3. Compare individual changes in Q to changes in depression severity.

Add third_party to the Matlab path for the code in functions to run.

1. Modularity estimation

Timecourses are extracted from a set of regions to measure functional connectivity as estimate normalized brain network modularity. eg:

% define imaging matrix
imaging_volume = 4D_brain_imaging_brain

% define atlas matrix
atlas_volume = 3D_regional_brain_atlas

% Estimate norm. modulatrity
Q_norm = calculate_static_modularity(imaging_volume, atlas_volume)

NB:- Imaging and atlas data need to be in the same space (e.g., MNI)

The modularity estimation was heavily inspired by Karolina Finc's Nature Communications paper on working memory: https://www.nature.com/articles/s41467-020-15631-z. To normalize Q, the maximum true Q is divided by the mean of a distribution of shuffled functional connectivity matrices.

2. Stats & plotting

The extracted Q and Beck Depression Inventory (BDI) measures are in .mat files here:

data/
|-- psilodep1/
|   |-- dat_1.mat
|
|-- psilodep2/
|   |-- dat_2.mat

BDI

To recreate the BDI figure that visualizes the reductions in depression severity following psilocybin therapy, run:

>>> BDI_figure

Imaging figure

To recreate the brain imaging figures that visualises reductions in modularity (Q) following psilocybin therapy, run:

>>> Imaging_figure

Stats

The additional stats reported in the paper are calculated in:

functions/
|-- psilodep1.m
|-- psilodep2.m

This code is no longer supported

After leaving academia, I no longer have access to Matlab licenses!

Get in touch if you:

• find a bug
• would like to translate this repo into an open source language
• want help with applying this sort of analysis to your data.