README.md

In anbai106/NeuroStatisticR:

This repository contains a software framework for mixd effect generalize linear model (GLM) for group comparison based on R. These software could be adapt to any atlas-based ROI measurement.

Citing this work

If you use or adapt this software, please cite:

Bertrand, A., Wen, J., Rinaldi, D., Houot, M., Sayah, S., Camuzat, A., Fournier, C., Fontanella, S., Routier, A., Couratier, P. and Pasquier, F., 2018. Early cognitive, structural, and microstructural changes in presymptomatic C9orf72 carriers younger than 40 years. JAMA neurology, 75(2), pp.236-245.- Paper in PDF

Wen, J., Zhang, H., Alexander, D.C., Durrleman, S., Routier, A., Rinaldi, D., Houot, M., Couratier, P., Hannequin, D., Pasquier, F. and Zhang, J., 2018. Neurite density is reduced in the presymptomatic phase of C9orf72 disease. J Neurol Neurosurg Psychiatry, pp.jnnp-2018.- Paper in PDF

Documentation

This code relies either on the Clinica software platform that you will need to install, or other any neuroimage-based analysis software, such as FreeSurfer or SPM. After preprocessed by these softwares, you have to organize the resulting ROI measurement into a tsv file with a standard format (You can use this script to extract the stats file for each image into a single tsv file: NeuroTsvWriter.

Example

Let's assume that you have run Clinica pipeline to extract the Desikan atlas mean ROI measurements for T1-weighted MRI, the output tsv should be organized like this:

participant_id  session_id  gender      group   age ROI 1   ROI N
sub-CLNC0001    ses-M00     Female  CN      71.1    107434  3155
sub-CLNC0002    ses-M00     Male    CN      81.3    117133  3211
sub-CLNC0003    ses-M00     Male    CN      75.4    103133  3092
sub-CLNC0004    ses-M00     Female  CN      73.9    117117  2999
sub-CLNC0005    ses-M00     Female  AD      64.1    117155  3184
sub-CLNC0006    ses-M00     Male    AD      80.1    127199  3088
sub-CLNC0007    ses-M00     Male    AD      78.3    112100  3309
sub-CLNC0008    ses-M00     Female  AD      73.2    133222  3122

Of note, here is just an example for the tsv file, you do not have to follow exactly the same order as here, the most important is to give correspondence for the columns' names and in the source code.

The model of interest is:

Y-ik(j) =μ+β×Gender-i +λ×Age-i +η×Group-i +U-k +Ε-ik(j)

where Yik(j) is the response of the j-th region of interest for the i-th participant and the k-th family; Gender-i, Age-i, and Group-i are the fixed effects; μ, β, λ, and η are their estimated parameters; U-k is the random effect measuring the difference between the average response in the family and in the whole population; and Ε-ik( j) is the random error.

Install this software in Rstudio

Imagine that you have Rstudio installed on your machine. You can install this software as a package from Rstudio. After cloning this repository at your local position: repo_path:

Run this command in your Rstudio console:

setwd("repo_path")
install("NeuroStatisticR")

Check the documentation of the pipeline:

?t1_stats_pipeline

Then it is your responsibility to adapt the framework into your specific question or model.

anbai106/NeuroStatisticR documentation built on May 27, 2020, 2:20 a.m.