November 2023#

This roadmap reflects our current priorities which are to refactor and organise the BrainGlobe codebase. Many of these changes won’t add any new features - they are largely aimed at making existing features more accessible (e.g. through a graphical user interface), ensure all tools are easier-to-install and at making BrainGlobe easier to maintain and further develop in the long-term. We hope this will ensure that BrainGlobe is around for years to come, and that we have a solid platform to develop new tools and support a growing community in the future.

Current status#

Issues#

  • brainrender has considerable technical debt (brainrender/247)

  • Lots of repositories makes it more difficult to understand which tool does what, and to contribute (e.g. cellfinder, cellfinder-core, cellfinder-napari)

  • The mix of napari plugins, Python APIs and command line tools make the ecosystem appear more complicated than it should

  • It is not yet possible to carry out all analyses and visualisation within a single platform (napari)

    • Some non-image processing steps need to be carried out separately (e.g. assigning cells to brain regions)

    • Some visualisation relies on exporting data from napari to brainrender

v1 (Q4 2023)#

  • brainrender installable alongside analysis tools (cellfinder, brainreg, brainreg-segment)

  • All tools installable via pip install brainglobe or conda install brainglobe -c conda-forge

v2 (Q2 2024)#

  • Consistent API, e.g. from brainglobe import cellfinder or from brainglobe import cell_detector_3D (or both)

  • Carry out everything in v1 without leaving napari

  • Automated benchmarks for analysis tools (cell detection & 3D atlas registration) on “real” (full size) data

  • Combine napari plugins with backend to reduce number of repositories (BrainGlobe/3 & (BrainGlobe/4)

  • Remove existing cellfinder CLI to simplify naming (BrainGlobe/6)

  • Release brainglobe-workflows, a collection of scripts to facilitate common tasks (including replacing the cellfinder CLI)

v3 (Q4 2024)#

  • Intuitive navigation of napari plugins

    • Create “metaplugin”

  • Generic layer interpreters (BrainGlobe/10)

  • Refactored and documented Python API

  • Facilitate common analyses/visualisation within napari

    • Plot cells per brain region

  • Sample data provided via napari

Other ongoing projects (release TBC)#