Importing data


DICOM data is the easiest to import. Currently MR, CT, SR, CR, US, and OT modalities are supported. Place the dicom files in the /nidb/dicomincoming directory. The DICOM files must be uniquely named and they must not be in sub directories.

Once the files are in place, the parsedicom module should pick up the new files, parse them, and archive them. It should match new data to existing subjects, or create new subjects if they do not exist. It matches based on Name, DOB, and Sex.

NiDB Data Packages

Data exported through the NiDB export process can be imported by placing the .tar.gz file into the /nidb/packageimport directory. The import module will automatically parse the contents of the file and add the data to the archive.

Existing Datasets

ABIDE – Provided by the INDI project, this project has collected imaging and phenotypic data for 1,112 subjects and made them available for anyone to download and analyze. It is possible to import this dataset into NIDB using a script. Follow these instructions to import ABIDE imaging and phenotype data into your installation of NIDB.

  1. Install NIDB on your own server, or download the virtual machine (VM).
  2. Install FSL on your server (FSL is already installed on the VM)
  3. Download the ABIDE data .tgz files to a directory such as /nidb/import/abide
  4. In the /nidb/programs directory run ./ /nidb/import/abide
  5. Perform the following command only once to import the phenotypic data:
  6. cd /var/www/html
    php parse_abide_phenotype.php
  7. Data should now be imported!
Notes: Because the actual birthdate of the subjects is unknown, a default DOB of July 4, 1776 is entered, however the age at scan is correctly entered for each imaging study and will appear on query results.

ABIDE Dataset list: Caltech, CMU_a, CMU_b, KKI, Leuven_1, Leuven_2, MaxMun_a, MaxMun_b, MaxMun_c, MaxMun_d, NYU, OHSU, Olin, Pitt, SBL, SDSU, Stanford, Trinity, UCLA_1, UCLA_2, USM, Yale


2 thoughts on “Importing data

  1. What different kinds of file types besides dicom can be supported? NIFTI? E-files? .Mag files too? Thanks!

    • Any type of file can be stored in the database, which includes single or multi-file, and associated behavioral files (eprime, etc), and any of those files can be uploaded manually. For importing large sets of data automatically, DICOM and PAR-REC can be imported in bulk by parsing the file’s header data. There also have import scripts for Nifti, Neuroscan EEG, Eye-tracking, and SNP (genetic) data. The meta data for those formats comes from the directory structure of the imported files or is specified in the script. So, for example a directory name might contain the ID of the subject, and the sub-directory names might contain the series description of the data. If you need help customizing an import script for your data, just let me know and we can probably get it done.

Leave a Reply to Pamela Ng Cancel reply

Your email address will not be published. Required fields are marked *


eight + = 16

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>