
Open-source tools for neuroscience and ML
Showing 5 of 5 projects. Click any project card for scope, mentors, and proposal studio.
Mentors: Student: Soumya Snigdha Kundu
Cellfinder currently expects full 3D whole-brain volumes with both signal and background channels, so brain slices (2.5D) and single-channel acquisitions are not first-class citizens. The codebase also mixes axis conventions, which makes dimensionality changes error-prone. A recent data-loading refactor introduces explicit axis tracking and a new class hierarchy for cube generation and augmentation; the 2D/2.5D and single-channel work proposed here will build on that foundation. The core deliverables are to standardise axis semantics as the foundation, add a 2D detection path for brain slices and refactor classification to support both 2D and 3D inputs, and make the background channel optional across the main API and classification pipeline.
Mentors: Student: Anna Teruel-Sanchis
This project aims to develop a napari-based GUI for interactive correction of pose-estimation tracks within the movement framework. It will enable users to resolve identity swaps in multi-animal datasets and correct keypoint errors efficiently. In multi-animal tracking, automated pipelines often fail due to occlusions, noise, and identity ambiguity. The proposed tool integrates visualization and editing into a unified workflow, allowing users to inspect and refine trajectories directly while maintaining standardized data structures. This will improve the quality and reliability of pose-estimation datasets, providing the open-source community with a robust and extensible solution for behavioral data curation. This project builds on an initial napari-based prototype that I have already developed for manual trajectory correction.
Mentors: Student: Varun Singh
This proposal aims to improve the infrastructure of the BrainGlobe ecosystem and make it easier to maintain and contribute to as it continues to grow. This will be done by simplifying the BrainGlobe meta-package API, aligning function and argument naming across tools, improving CI workflows by migrating from tox to uv, automating API documentation, and refining the conda-forge release process. Deliverables: Expose a simplified API from individual BrainGlobe tools via the meta-package. Standardise naming, arguments, and structure across tools. Migrate CI workflows from tox to uv. Improve conda-forge packaging and release setup. Automate API documentation and publishing, clean up badges and project metadata across repos.
Mentors: Student: Saarah Hussain
Synopsis: This project aims to enhance the brainglobe-registration tool by automating the selection of atlas regions for aligning 2D or 3D brain images to a standard anatomical reference via brainglobe-atlasapi. Currently, this process is manual, time-consuming, and subjective, posing a barrier to reproducibility and efficiency. By introducing a new preprocessing step that automatically identifies the optimal atlas region – using methods such as adaptive grid search, machine learning, or Bayesian optimisation – this project will streamline the workflow and improve accuracy. Deliverables include at least two implemented approaches, benchmarking, comprehensive testing, full documentation, and a blog post to showcase the new functionality. This automation directly benefits my research in paediatric radiotherapy, where mapping histology images to an atlas is essential for analysing cellular behaviour across brain regions. By making this process faster and more robust, it not only enhances my own work but also enables broader adoption of BrainGlobe tools by researchers in neuroscience and healthcare, even those without extensive anatomical expertise. Minimal Set of Deliverables: • Implement a new preprocessing step to automatically select the atlas region. • Compare at least two region-selection methods (e.g., adaptive grid search vs ML/Bayesian optimisation). • Add tests to validate the new functionality. • Document the new pipeline and add usage instructions. • Write a blog post showcasing the project and results.
Mentors: Student: Shrey Singh
This is a proposal to extend Datashuttle's functionality to allow connections to Google Drive and AWS buckets as remote storage options. This project aims to improves the flexibility and adoption of Datashuttle to allow neuroscience researchers use a variety of storage options with the most popular ones being Google Drive and AWS buckets. This will be done by extending Rclone's wrapping of Datashuttle and extending the terminal user interface to allow connections to Google Drive and AWS buckets. **Deliverables** - Support for setting up connections to Google Drive and AWS and transferring data in the Datashuttle's Python API using Rclone. - TUI tabs and screens to enable users to setup connections to Google Drive and AWS and transfer data via the TUI. - Integration and TUI tests for the implemented features. - Documentation for setting up Google Drive or AWS as central storage.