This documentation is related to the Cytomine ULiège Research & Development (opens new window) Edition from the research team at Montefiore Institute, University of Liège, Belgium. It includes the Cytomine Community Edition (opens new window) features plus latest experimental features driven by our research projects and collaborations but that are not yet fully validated (therefore not yet in the official Cytomine version). See below for examples of such experimental features.
For imaging scientists and biomedical researchers
We are extending the official Cytomine version for cutting-edge imaging modalities (beyond histology incl. 2D+c+z+t, hyperspectral data, spatial proteomics, imaging mass spectrometry, ...) and their combination using Image Groups and Annotation Links for multimodal studies.
For data and computer scientists
We are improving the Cytomine software execution architecture to support multiple types of algorithms and execution environments, and to perform benchmarking (see Biaflows project). We are also extending Cytomine internal data models and its RESTful API to support various types of data import/export.
In addition to standard features of the official Cytomine version for online courses, at ULiège we developed learning analytics modules (in Python) to help teachers monitor student's activities in Cytomine-based histology courses.
For bioimage analysts and pathologists
We are developing novel AI algorithms to ease bioimage analysts and pathologists to analyze their data including interactive annotation tools and algorithms for various quantification tasks (cell counting, tumor delineation, content-based image retrieval,...)(see ULiège R&D publications).
For other (non-biomedical) researchers
Through collaborations, our team is applying and extending some Cytomine features on large image sets from other fields such as digital collections of works of art, industrial quality control, remote sensing, ...