Goals to aim for during Diane's visit.

• Abstract segmentation into a general structure to enable plug-and-play of multiple deep learning models (Diane/Alán/Peter).
• Integrate interface changes into the pipeline (Alán).
• Document the process of setting up both user and development environments, and running/analysing standard datasets (Ivan/Francois).
• Identify experiments with fringe cases (e.g., different types of metadata, time-points and channels). (Arin/Ivan/Francois).
• Add tests for metadata, and the Image and Dataset classes (Ivan/Francois).
• Add docstrings for classes and methods, prioritising the most used sections (Peter).
• Implement test coverage into CI (Arin/Alán).
• Identify bugs in untested experiments and document them (Ivan/Francois).