Skip to content
forked from aces/Loris-MRI

The set of scripts that preprocess and insert MRI data into the database.

Notifications You must be signed in to change notification settings

cmadjar/Loris-MRI

 
 

Repository files navigation

This Readme covers release 24.1.* of the LORIS Imaging Insertion Pipeline for Ubuntu or CentOS systems

This repo accompanies the LORIS neuroimaging data platform main repo, release 24.1.*.
For documentation and detailed setup information, please see the LORIS-MRI documentation for your installed version.

This repo can be installed on the same VM as the main LORIS codebase, or on a different machine such as a designated fileserver where large imaging filesets are to be stored.

System Requirements

On Ubuntu, DICOM toolkit will be installed by the imaging install script (step 4 below). This script will apt-get install dcmtk.

For CentOS: The LORIS wiki Imaging setup page (see Section 1, installing codebase) includes links to older transcripts for CentOS installation and notes on dependencies including DICOM toolkit.

The following installation should be run by the $lorisadmin user. sudo permission is required. See aces/Loris README.md for further information.

Installation

1. Create directories and download Loris-MRI code

sudo mkdir -p /data/$projectname
sudo mkdir -p /opt/$projectname/bin/mri
sudo chown -R lorisadmin:lorisadmin /data/$projectname
sudo chown -R lorisadmin:lorisadmin /opt/$projectnawe
cd /opt/$projectname/bin

Get the code: Download the latest release from the releases page and extract it to /opt/$projectname/bin/mri

2. Install Python 3 with pip and virtualenv

sudo apt install python3 
sudo apt install python3-dev
sudo apt install python3-pip
sudo apt install libmysqlclient-dev
sudo apt install virtualenv

3. Install MINC toolkit from http://bic-mni.github.io/

Download the pre-compiled package for your operating system. Install required dependencies such as imagemagick. Then install your MINC toolkit package:

sudo dpkg -i minc-toolkit<version>.deb

Then source the MINC toolkit environment by running (for bash) source $mincToolsDirectory/minc-toolkit-config.sh or (tcsh) source $mincToolsDirectory/minc-toolkit-config.csh,

where $mincToolsDirectory is the path where the MINC toolkit is installed (e.g. /opt/minc/ OR /opt/minc/$mincToolsVersion/ for more recent installs)

For the defacing scripts, you will also need to download the pre-compiled bic-mni-models and beast data and model packages for you operation system.

sudo dpkg -i bic-mni-models-<version>.deb
sudo dpkg -i beast-library-<version>.deb

4. Run installer to set up directories, configure environment, install Perl libraries and DICOM toolkit:

cd /opt/$projectname/bin/mri/
bash ./imaging_install.sh

You will be asked for the following input:

  • What is the database name? $dbname
  • What is the database host? $dbhost
  • What is the MySQL user? $lorisuser [Use the same mysql user from the Loris installation, i.e. lorisuser]
  • What is the MySQL password?
  • What is the Linux user which the installation will be based on? $lorisadmin
  • What is the project name? $projectname
  • What is your email address?
  • What prod file name would you like to use? default: prod [leave blank]

If the imaging install script reports errors in creating directories (due to /data/ mount permissions), review and manually execute mkdir/chmod/chown commands starting at imaging_install.sh:L97

Note: The installer will allow Apache to write to the /data/ and /opt/ directories by adding user lorisadmin to the Apache linux group. To ensure this change takes effect, log out and log back into your terminal session before running the imaging pipeline. The installer will also set Apache group ownership of certain /data/ and /opt/ subdirectories.

5. Configure paths and environment

Ensure that /home/lorisadmin/.bashrc includes the statement:

source /opt/$projectname/bin/mri/environment

Then source the .bashrc file.

INSTALLATION COMPLETE!

Please refer to the Install section in the LORIS-MRI documentation for your installed version for:

  • customizations and protocol configurations (Section 2.2).
  • verifying that certain fields were correctly populated by imaging_install.sh (Section 2.3).

About

The set of scripts that preprocess and insert MRI data into the database.

Resources

Stars

Watchers

Forks

Packages

No packages published

Languages

  • Perl 75.6%
  • Python 23.1%
  • Other 1.3%