Type: Software,

Keywords: BIDS, Nifti, PET, Imaging, ECAT

PET2BIDS is a tool to help convert PET imaging data into the Brain Imaging Data Structure (BIDS)

This repository hosts tools to curate PET brain data using the Brain Imaging Data Structure Specification. The work to create these tools is funded by Novo Nordisk Fonden (NNF20OC0063277) and the BRAIN initiative (MH002977-01). This tool extracts and outputs some of the additional blood and radiological data along with PET neuroimaging data into the Brain Imaging Data Structure. It is one of the few tools that exists to aid in this conversion and is being used to convert and share data on

* Inject PET information at time of image conversion from Dicom/ecat to ease sharing/transmissibility of imaging and blood data via conversion to BIDS w/ this tool
* Python, Matlab, and a stand alone CLI provide flexibility in use among different environments

This tool is primarily aimed at aiding researchers in converting scanner data (post reconstruction) along with blood data into the Brain Imaging Data Structure (BIDS) for later sharing, analysis, and use. Once PET data is transformed into BIDS it can be combined and analyzed more easily as each different data set (multiple sites, multiple studies, etc.) shares a common structure. This allows for the reuse of data, analysis tools, and the de-duplication of effort on the part of researchers. The aim is to develop and provide this tool as a first step towards sharing data on open platforms, such as, to facilitate data sharing and reusability by easing the tedious of process of curating PET data.

* Convert ECAT or Dicom data into BIDS PET
* Collect additional metadata via text file/spreadsheet during conversion of images and into tsv formats
* Extract PET information from lesser known format ECAT via CLI

* Convert data while the experiment is fresh in the minds of researchers, capturing relevant information before it becomes lost
* Ease the burden on searching for or inspecting singular files to collect relevant information for BIDS/PET
* Uses tried and tested dependencies (Dcm2niix, Nibabel, Matlab, etc.) to provide the best fidelity during the conversion process

* Requires PET knowledge for best use
* Handling of heterogenous data is only as good as the data and the knowledge of the user, at least during first setup
* Command line/programmatic use may be a barrier to some users more comfortable with a graphical user interface

* Python, Matlab, and Dcm2niix must be installed to use this software


Anthony Galassi, Computer Programmer


National Institutes of Mental Health, Bethesda MD



Adam Thomas, Director of Data Science and Sharing Team, NIH
Anthony Galassi, Computer Programmer, NIH
Chris Rorden, Prof. Endowed Chair of Neuroimaging, U. of. South Carolina Dept. of Psychology
Cyril Pernet, Senior Research Software Developer, Neurobiology Research Unit
Gabriel Gonzalez-Escamilla, Post Doc.University Medical Centre, Johannes Gutenberg University Mainz
Martin Nørgaard, Post Doc., Neurobiology Research Unit – Stanford
Melanie Ganz, Assistant Professor, Neurobiology Research Unit UCPH
Remi Gau, Researcher, Université catholique de Louvain, Crossmodal Perception & Plasticity lab



* Novo Nordisk fonden NNF20OC0063277
* BRAIN initiative MH002977-01