Source Space and Forward_Solutions
- Define cortical surface source spaces
a. Use standard steps (incl. patch decompositions) outlined in MNE manual
b. Same as used by SPIGH code
- Define subcortical volume source spaces:
a. In Shell
i. Reconstruct MRIs via Freesurfer Stream
ii. Setup a standard subject folder based on MNE/Freesurfer manual instructions
iii. This will have all info to extract volumes of major subcortical structures
b. In MATLAB: Create .mgz, .txt and .mat files with masks, spatial coordinates and geometric information for each subcortical region’s source spaces. Run exec_srcfwd_final. This calls
i. read_save_sfwd to perform dimensionality reduction, split into patches/subvolumes
ii. check_subdivs to create mgz files for each subdivision in mri folder
iii. srcspace_figure to generate mgz file illustrating subdivided source space figure
- This codestream will lead you to run shell script functions as needed too:
a. Generate cortical forward solutions using standard steps from MNE manual
b. Run scort_srcfwd_exec for subcortical volumes, hippocampus and brainstem
- Scripts to generate figures in the paper are under \FigureS1
- In Freeview: view .mgz, label and surface files together to check if all patch/subvolume/surface source spaces are done properly, so as to be sure what regions forward solutions correspond to
- In Freeview, visualize subcortical source space masks to ensure all done correctly
Data_Process: Remove artifacts, Filter Data; Evoked Averages; Compute Covariances
MATLAB code to read data, process source/noise covariances, define parameter choices, obtain MNE estimates, obtain SPIGH estimates for given hierarchy, transition across hierarchies
- Scripts to run cortical inverse solution (within and across all cortical hierarchies)
- Scripts to visualize cortical inverse solution (time series, spatiotemporal movies, performance metrics)
- Scripts to run subcortical inverse solution (with reduced cortical source space and all deep sources)
- Scripts to visualize end joint inverse solution (time series, spatiotemporal movies, performance metrics)
Note: running this code will display instructions on shell commands to run as and when required.
Core_Functions: General Sub-Functions Called for Angles, Field Maps, Algorithm, Visualization, ETC
Sparsity enables subcortical source estimation
Pavitra Krishnaswamy, Gabriel Obregon-Henao, Jyrki Ahveninen, Sheraz Khan, Behtash Babadi, Juan Eugenio Iglesias, Matti S. Hämäläinen, Patrick L. Purdon, Proceedings of the National Academy of Sciences Nov 2017, 114 (48) E10465-E10474; DOI: 10.1073/pnas.1705414114
https://www.pnas.org/content/114/48/E10465.full