gVXR + CIL + qim3d: Training and Mini-Hackathon – AIDA

Tomographic imaging remains state-of-the-art for non-invasively mapping internal microstructure in 3 or more dimensions with applications in a variety of fields. Lots of developments are happening in new experimental setups at both laboratory and large facilities such as MAX IV and ESS allowing faster and higher-resolution X-ray and neutron imaging as well as time-resolved, energy-resolved, multi-modal and in situ experiments with incomplete data. Researchers need advanced computational tools for the different steps of the imaging pipeline.

In this 3-day event you will get hands-on experience with three open-source libraries for simulating, reconstructing and analyzing tomography data.

On day 1, the focus is simulation using gVXR (https://gvirtualxray.sourceforge.io/) – virtual X-ray imaging library on GPU given by lead developer Franck Vidal.

On day 2, we focus on reconstruction using the Core Imaging Library (CIL, https://www.ccpi.ac.uk/CIL/) with introduction and hands-on exercises by two of the core developers Jakob Sauer Jørgensen (DTU Compute & QIM Center) and Gemma Fardell (STFC, UKRI).

On day 3, we cover image analysis such as segmentation and quantification using the qim3d (https://docs.qim.dk/qim3d/) library, by the lead developer Felipe Delestro (DTU Compute & QIM Center).

The event is open to beginners as well as participants with experience. Mornings will be dedicated to introductory lectures and hands-on exercises. Afternoons will be more open-ended with the possibility to explore more advanced topics as well as apply the libraries to own data. Example data cases will be available for exploration with the three libraries.

Participants are encouraged to bring a poster for our informal Posters & Pizza event on the first evening to facilitate collaboration and knowledge sharing.