BIDS apps: Improving ease of use, accessibility, and reproducibility of neuroimaging data analysis methods

Abstract

The rate of progress in human neurosciences is limited by the inability to easily apply a wide range of analysis methods to the plethora of different datasets acquired in labs around the world. In this work, we introduce a framework for creating, testing, versioning and archiving portable applications for analyzing neuroimaging data organized and described in compliance with the Brain Imaging Data Structure (BIDS). The portability of these applications (BIDS Apps) is achieved by using container technologies that encapsulate all binary and other dependencies in one convenient package. BIDS Apps run on all three major operating systems with no need for complex setup and configuration and thanks to the comprehensiveness of the BIDS standard they require little manual user input. Previous containerized data processing solutions were limited to single user environments and not compatible with most multi-tenant High Performance Computing systems. BIDS Apps overcome this limitation by taking advantage of the Singularity container technology. As a proof of concept, this work is accompanied by 22 ready to use BIDS Apps, packaging a diverse set of commonly used neuroimaging algorithms.

Fig 1. Overview of the creation and use of BIDS Apps.

The source code of each App is stored in separate GitHub repository. Each repository is connected to a Continuous Integration server responsible for building testing and deploying the corresponding App. For every new release of an App, a new container image is deposited in Docker Hub. Users can directly download and run the BIDS Apps container images either directly using Docker on any Windows, Mac, and Linux machine or convert them to Singularity and run them on an HPC.

Fig 2. The overall structure of workflows.

Workflows will need to decouple the individual level analysis (process independent subjects) from the group-level analysis. For the analysis of individual subjects, the workflow will require an understanding of the BIDS structure so that the required inputs for the designated subject are found. The optional reducer module will take up from results generated in the mapper and generate a group output. The overall workflow has an entrypoint and an endpoint responsible of setting-up the map-reduce tasks and the tear-down including organizing the outputs for its archiving, respectively. Please mind that each app can implement multiple map and reduce steps (see Advanced use cases).