Modeling conduction from g-ratio

This repository includes the code and dara required to reporoduce the figure in: "Modeling conduction delays in the corpus callosum using MRI-measured g-ratio" Berman S., Filo S., Mezer A.A.. NeuroImage (2019).

The pipeline and methods used inthe paper are described below.

Analysis steps and requirements

For modelling the conduction callosal tracts using g-ratio, we need:

1. g-ratio maps

We calculated g-ratio using:

1. MTV as MVF, calculated with mrQ [https://github.com/mezera/mrQ])
2. NODDI parameters, calculated using AMICO [https://github.com/daducci/AMICO]
3. The formula for the calculation as described in: Stikov, N., Campbell, J. S., Stroh, T., Lavelée, M., Frey, S., Novek, J., ... & Leppert, I. R. (2015). In vivo histology of the myelin g-ratio with magnetic resonance imaging. Neuroimage, 118, 397-405.‏

FVF = MVF + (1-MVF)(1-Viso)Vic

g = sqrt(1-(MVF/FVF)).

Examples of (slices of) g-ratio maps can be seen ../MR_ConductionModeling/Figures/Supp_Fig1_maps.m

2. Corpus Callosum segmentation

1. We performed whole-brain probablistic tractography using mrTrix's anatomicallly constrained tractography [https://mrtrix.readthedocs.io/en/latest/quantitative_structural_connectivity/act.html]. We use the command: ['/opt/mrtrix3/bin/tckgen ' csdFile ' ' tckNew ' -act ' t1w5ttOUT ' -seed_gmwmi ' GWinterface ' -backtrack -cutoff 0.1 -crop_at_gmwmi -select ' tracksNum ' -maxlength ', '200']

2. Then we used AFQ to segment the callosal tracts.

   • First run dtiInit on the difusion data: [https://github.com/vistalab/vistasoft/tree/master/mrDiffusion/dtiInit]
   • Then run AFQ: [https://github.com/yeatmanlab/AFQ]. AFQ will take as input
     • The ACT-generated tracography
     • The g-ratio map.
     • The dtiInit-generated dt6.mat file.
   • The AFQ will save the separate callosal tract as fg files.

3. Finally we reSaved the fg files after sampling g-ratio per streamline. See an exmaple for sampling gratio in the white matter in ../MR_ConductionModeling/Analysis/Sample_g_along_strmlns.m

Examples of the sampled tracts can be found in ../MR_ConductionModeling/Figures/Fig2_CCfibers.m

3. Simulation of action potential conduction along white matter

We use the simulation imlemented by David Attwell's lab: [https://github.com/AttwellLab/MyelinatedAxonModel]

An example of our use of this code can be found in ../MR_ConductionModeling/Analysis/RunModel.m

4. Simulation of LFP signal from input latency

We use the simulation written by Hermes, Nguyen and Winawer (2017, PLOS Biology)

The original code is available in: [https://github.com/WinawerLab/BOLD_LFP]

For our adaptation of a small part of the code, used to derive LFP from a distribution of latencies (over streamline) see: ../MR_ConductionModeling/Analysis/simLFP.m

Shai Berman, Mezer Lab, 2019 (c)