If this code is used in a publication, please cite the manuscript:
    Huang, H., Gregg, N. M., Valencia, G. O., Brinkmann, B. H., Lundstrom, B. N., Worrell, G. A., ... & Hermes, D. (2023).
    Electrical stimulation of temporal and limbic circuitry produces distinct responses in human ventral temporal cortex. Journal of Neuroscience.
    DOI: https://doi.org/10.1523/JNEUROSCI.1325-22.2023
    
Contact info: Harvey Huang, [email protected]

Updated 2023/04/20 after second JNeurosci resubmission

*****

DEPENDENCIES

- SPM12: https://www.fil.ion.ucl.ac.uk/spm/software/spm12/
- mnl_ieegBasics: https://github.com/MultimodalNeuroimagingLab/mnl_ieegBasics
- matmef: https://github.com/MultimodalNeuroimagingLab/matmef
- mnl_seegview: https://github.com/MultimodalNeuroimagingLab/mnl_seegview
- vistasoft: https://github.com/vistalab/vistasoft
- Freesurfer v7: https://surfer.nmr.mgh.harvard.edu/fswiki/DownloadAndInstall
- GIfTI: https://github.com/gllmflndn/gifti

*****

USAGE

i. The data used in this analysis will be available upon manuscript publication on OpenNeuro, in BIDS format, at https://openneuro.org/datasets/ds004457.
	Please download all the data and copy into a folder named "data", in the root level code directory.

ii. Pial, cortical, and subcortical segmentations for each subject were obtained using Freesurfer v7. The relevant Freesurfer outputs for each subject are located in data/derivatives/freesurfer


Step-by-step script sections to generate all outputs are in "main.m". Please open "main.m" and follow the instructions below.
--

1. Calculate MNI 152 coordinates for electrodes in subjects 1 through 4
	ia. Use SPM12 to generate MNI152 forward deformation field from each subject's T1-weighted nifti
		Press Batch. > Spatial -> Segment.
		Specify the subject's T1w nifti file for the "Volumes" input. Select "Forward" for "Deformation Fields". Press the play button.
		Keep the file titled "y_[t1FileName].nii". all other output files can be deleted
	This MNI 152 forward transformation was performed using the non-defaced AC-PC space T1w MRI as input for subject 1 and using the non-defaced native space T1w MRIs as inputs for subjects 2 through 4. Subject 5 was not used in consensus BPC analysis, and therefore not transformed to MNI 152 space.
	ib. Run saveMNIelectrodes.m. Outputs are saved to 'output\sub-X\MNI152\.'

2. Values for Table 1. Run determineStimSites.m to calculate the number of stimulation sites per subject (including stimulated measurement electrodes)

3. Figure 1C-E: Projection matrix, significance matrix, and NNMF illustrations. Run methodfigMatrices.m

4a. Figures 2A-D and 6: Save subject-level BPCs and per-subject wavelet spectrogram, broadband estimates for all measurement electrodes. This will take a while to run
- Run saveBPCs.m

4b. Figures 2E-F: To save the T1 MRI slices and inflated pial renderings for subject 1
- Run plotInflatedT1slicesSub1.m Manually click on coronal view to get sagittal and axial views at (z, x) = (-15.0, -23.0) prior to running the last section.

5. Save cortical labels to stim sites in *_BPCSNR.tsv outputs for all electrodes/subjects
- Run assignStimDestrieux.m
- LA1-LA2 stim sites in sub-1 measurement electrodes were manually corrected (in the TSV) afterward to "Left_Amygdala" by visual inspection of CT/T1w MRI overlay

6. Figure 3 and Extended Data Figure 3-1. Calculate consensus BPCs and plot anatomical distributions on MNI 152 brain. Also calculates cross-validation accuracy of consensus BPC model
- Run clusterBPCs.m
- For cortical and subcortical anatomical legends in Figure 3D (right), run bpcRegions.m

7. Figure 4. Plot stim sites by consensus BPCs on slices of MNI152 T1 MRI
- Run plotBPCs2MNISlices.m, manually adjusting z-coordinate by values on line 38

8. Figure 5. Calculate spectrograms and broadband estimates for consensus BPCs, across collateral sulcus measurement electrodes in subjects 1-4
- Run fig5_exStimSite.m to get the stim site in figure 5A
- Run bpcSpectraBB.m
- Panel A corresponds to RC2-3 stim pair recorded at electrode 1 in subject 2

9. Figure 7. Create Basis profile spectrogram (BPS) outputs and Adjusted Rand Index outputs quantifying similarity to subject BPC outputs
- Run overlapWithBPS

Extended Data Figures, Supplement
--

10a. Figure 2-1, panels A-D. Save monopolar subject-level BPC outputs without common average rereferencing.
- Run BPCsNoCAR
- To plot example of channels that go into construction of the common average reference, run figRev1_CARex.m. (not in manuscript)

10b. Figures 2-1, panels E-F: To save the T1 MRI slices and inflated pial renderings for subject 1 in the absence of re-referencing
- Run plotInflatedT1slicesSub1_noCAR.m Manually click on coronal view to get sagittal and axial views at (z, x) = (-15.0, -23.0) prior to running the last section.

11. Figure 3-2: Analysis of distance between stimulation site and measurement electrode vs. consensus BPC assignment or BPC SNR
- Run distanceVBPC.m

12. Figure 3-3: Calculate alternative single-step BPCs by pooling CCEPs across all subjects and applying BPC algorithm
- Run consensusBPCsPooled.m

13. Figure 4-1 (BPCs only) and all of Figure 4-2. Plot stim sites, colored by consensus BPCs, on subject inflated pial surfaces; Also plot BPCs colored and ordered by consensus identity
- Run plotInflatedBPCs.m

14. Figure 4-2 MRI slices. Plot stim sites by consensus BPC color on subject T1 MRI slices
- Run plotBPCs2Slice.m. Manually change sub and ch variables and click within hippocampus on coronal view. The (z, x) coordinates plotted for subjects 1 through 4 are (-15.0, -23.0), (-15.0, 23.0), (-13.0, -24.0), and (-13.0, 26.5), respectively.

15. Figure Response-1 in second resubmission. Plot subject BPC curves and anatomical distributions in each subject, in the absence of the exponential weighting/unweighting steps
- Run figRev_noExpWeight.m
- Type in zeta threshold when prompted. For panels A-C, zeta = 1. For panels D-F, zeta > 1.31 (e.g., 1.4)

16. Figure QC1 (quality check figure in OpenNeuro). Comparison of line noise removal methods on smearing of stimulation artifact: spectrum interpolation method used in this paper vs Butterworth notch filter
- Run figQC1_LNremoval

17. Figure QC2 (quality check figure in OpenNeuro). Validity of the BPS analysis, that the BPC algorithm is spatio-temporally invariant and can be applied to flattened 2D inputs.
- Run BPSshuffle