Use Case 25 – WebKnossos, a web-based tool for 3D data viewing and annotation

WebKnossos, a web-based tool for 3D data viewing and annotation, as a remote service.

webKnossos will be used to provide users with an intelligent tool for viewing and annotating neuroscience data for establishing the ground truth for machine learning algorithms.

Main Contributors

• A. Pacureanu – Lead scientist for neuroscience at ESRF (ERC grant 852455)
• T. Vincent – Graphics specialist at ESRF for visualisation
• F. Sturma – IT support for cloud infrastructure at ESRF
• N. Rzepka – Main contact for webKnossos at scalable minds

Use Case Action Flow

1. Deploy: Install webknossos
2. Import data: Import neuronal data into webknossos database and format
3. Test with users: Test with real users

Description of needs

The service needs 50 TB for neuroscience data.

Impacts from the implementation

The webKnossos service will provide users with a remote service for viewing and annotating large 3D datasets (Terabytes) without having to download the data nor install the software locally. Collaborative annotation of data will help scientists establish the ground truth for ML algorithms. webKnossos will be extended to deal with new applications, e.g. data from the Human Organs Atlas.

Other partners interested in this use case

• Scalable minds
• University College of London

Resources

• https://github.com/scalableminds/webknossos
• webknossos website
• Boergens, K., Berning, M., Bocklisch, T. et al. webKnossos: efficient online 3D data annotation for connectomics. Nat Methods 14, 691–694 (2017). https://doi.org/10.1038/nmeth.4331
• Neural cartography at ESRF

Contact person

Andy Götz (ESRF)