Longitudinal processing pipeline

This pipeline has some scripts to use recon-all and extract stats.

Main reference: http://freesurfer.net/fswiki/LongitudinalProcessing

Getting Started

Software requirements

• freesurfer
• MATLAB
• Sun Grid Engine
• LobesScalingCode, MATLAB toolbox

Steps

1. First, export your data using BIDS specs.

2. Copy file .env_sample renaming to .env and set the environment variables inside this file with the path for each directory.

3. Run recon-all using this command in a terminal:

   ./01-preproc_sge.sh

   This script looks for each session present in the BIDS dataset and performs the command recon-all for them. Output directories that already exist are skipped. It also generates the template and longitudinal folders. Reference: http://freesurfer.net/fswiki/LongitudinalProcessing

4. recon-all again, but this time with the flag -localGI. Run this command in a terminal:

   ./02-scriptalllgi.sh

   This step requires the software MATLAB. Output are located at $LOGDIR.

   Note: Check for errors using the script check-logs.sh

5. Extract some stats using this command in a terminal:

   ./03-extracts_stats.sh

   The results will be placed inside $STATSDIR

6. In MATLAB, run this command:

   lobes_scaling_code.m

   The results will be placed inside the output directory of each subject, inside the directory LobesScaling