March 2026#

Recent progress#

We’ve created three new software tools, and we’ve worked on improving and maintaining the BrainGlobe infrastructure. Specifically we have made

  • An alpha version of brainglobe-registration, allowing registration of 2D data (e.g. conventional tissue sections) to an atlas.

  • An alpha version of brainglobe-stitch, allowing efficient stitching of light-sheet data with a GUI.

  • An alpha version of template building tools, which have been used with help of the BrainGlobe team to make three new atlases.

  • An updated Pytorch backend for cellfinder, simplifying installation and support, and laying the groundwork for new analyses.

  • Improvements to test coverage, benchmarking tools, documentation and ease of installation.

  • Improvements and bug fixes to BrainGlobe atlases and related validation tools.

  • More atlases available via the BrainGlobe Atlas API.

High-level priorities for the next year#

Overall, the team will focus on further expanding access to BrainGlobe through new features, atlases and atlas-registered data. In particular, we are working towards

  • Enabling users to detect and analyse more cells, e.g. from c-Fos experiments.

  • Enabling users to detect and analyse cells from larger (e.g. rat) brains.

  • Enabling users to create their own templates and annotations and use them within BrainGlobe.

  • Making atlas downloads faster and leaner (BrainGlobe Atlas API v2).

  • Making diverse atlas-registered data such as gene expression maps accessible via BrainGlobe GUIs (BrainGlobe Data API).

  • Allowing easy registration of single histological sections and images of brain subvolumes to BrainGlobe atlases.

  • Adding existing API documentation to website.

  • Further expanding the number of atlases available via the BrainGlobe Atlas API.

Future plans#

The items below are in our plans, but not currently considered an immediate priority

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

  • A napari meta-plugin