Source code for research article:
"Differential auditory and visual phase-locking are observed during audio-visual benefit and silent lip-reading for speech perception"

by Máté Aller, Heidi Solberg Økland, Lucy J. MacGregor, Helen Blank, and Matthew H. Davis

Published in Journal of Neuroscience

Cite: Aller, M., Solberg Økland, H., MacGregor, L. J., Blank, H., & Davis, M. H. (2022). Differential auditory and visual phase-locking are observed during audio-visual benefit and silent lip-reading for speech perception. Journal of Neuroscience. https://doi.org/10.1523/JNEUROSCI.2476-21.2022

Installation

1. Clone or download this repository
2. Download all dependencies into their individual subfolders within audiovisual_speech_MEG/code/toolbox/
3. Download the data folder from here into the project root folder (audiovisual_speech_MEG/)
4. Build conda environment for python by executing:

   conda env create -f environment.yml
   

Usage with MATLAB

1. Navigate to the project root folder (audiovisual_speech_MEG/) and edit AVSM_setupdir.m adding the path to the project root folder on your computer. Then execute AVSM_setupdir. This takes care of setting up the environment.
2. Execute the matlab scripts found in results/ to reproduce behavioural and sensor space MEG results.
3. All analysis code can be found in audiovisual_speech_MEG/code/analysis/

Usage with Python

1. Setup environment by executing the commands below

   cd path/to/project/folder
   conda activate avsm
   setenv MESA_GL_VERSION_OVERRIDE 3.3
   setenv PYTHONPATH `pwd`
   

2. Update the project_dir variable in audiovisual_speech_MEG/code/analysis/megcoherence_utils.py to the project folder path on your system.
3. To reproduce the source-space MEG results run the audiovisual_speech_MEG/results/megcoherence_anatomical_roi_analysis.ipynb notebook in jupyter notebook.
4. Source analysis pipeline can be found at audiovisual_speech_MEG/code/analysis/megcoherence_source_analysis_pipeline.txt. To reproduce source coherence maps run the corresponding line of code from the text file in an Ipython console.

Dependencies (MATLAB)

• barwitherr
• bayesFactor
• Colormaps
• crameri
• fieldtrip
• gramm
• randpermfull
• shadedErrorBar
• suplabel

Dependencies (Python)

All python dependencies are listed in audiovisual_speech_MEG/environment.yml

Contributors

The MEG analysis code was written by Máté Aller. Preprocessing scripts (MEG, behavioural) were written by Heidi Solberg Økland.