Repository of code used for stratification of pre-HD by rate of cognitive decline using single visit resting state fMRI
Scripts used for analysis of resting state fMRI data in premanifest Huntington's Disease in: "Resting-state connectivity stratifies premanifest Huntington's disease by cognitive decline rate" Authors: Pablo Polosecki, Eduardo Castro, Irina Rish, Dorian Pustina, John H. Warner, Andrew Wood, Cristina Sampaio and Guillermo A. Cecchi
This is NOT a toolbox. The scripts are provided "as is" for the purposes of reproducibility and transparency. If interested in adapting these scripts, feel free to contact Pablo Polosecki (pipolose@us.ibm.com) for assistance.
Most scripts are written in Python (2.7), except for the ones in fcd_tools, which are written in MATLAB This repository contains the following folders (each folder has its own README file with a description of its contents):
-
explore: Scripts for manipulating files from the raw dataset, and obtaining longitudinal slopes of cognitive change.
-
preprocess: Scripts for preprocessing fMRI time series and performing registration to the MNI template.
-
fcd_tools: Scripts for computing FCD feature maps.
-
polyML: Scripts used for cross-validated classifications.
-
my_nipype_io: Modification of a nipype's io interface to deal with modernly formatted strings.
-
plotting_tools: Functions for plotting tasteful statistical maps on MNI space.
-
figure_scripts: Scripts used for producing brain map figures in the paper.
test_retest_correlations: Ipython notebook computing similarity between test and retest FCD maps.
-
controls: Scripts for controling for confounds reported in the paper.
controls/combat_harmonization: Corrects FCD features by site using combat
controls/combat_harmonization: Corrects FCD features by site using combat.
controls/VBM_controls: Corrects FCD features by gray matter concentration. Assumes VBM maps exist. (In our case, these were provided by Castro et al. 2018)
controls/n_visit_controls: Controls for effect of number of visits in assigment to extreme subgroups of decline
controls/site_effects: Checks for differences in demographics/cognition across sites.
controls/exclusion_numbers: Ipython notebook computing how many subjects where excluded for what reason
Dependencies The following softwares and libraries are required for running the scripts: Python 2.7, Pandas, Numpy, Matplotlib, Seaborn, Scickit-learn, Scikit-contrib Lightning, Nilearn, Nipy, Nipype, FSL, FreeSurfer, MATLAB
References Castro, E., Polosecki, P., Rish, I., Pustina, D., Warner, J. H., Wood, A., et al. (2018). Baseline multimodal information predicts future motor impairment in premanifest Huntington's disease. NeuroImage: Clinical, 19, 443?453. http://doi.org/10.1016/j.nicl.2018.05.008