The project Transcend focuses on the segmentation and analysis of ultrastructural electron microscopy (EM) images to study subcellular architecture using artificial intelligence methods.
It combines both assisted and automated segmentation of organelles and macromolecular complexes, enabling high-throughput characterization of cellular structures. Spatial and morphological features are then extracted and engineered to quantify patterns in cellular organization. These features serve as biomarkers for comparative analyses, supporting applications in structural biology, pathology, and computational microscopy.
While the original project centered on EM imaging, Transcend has since been successfully applied to radiological and trauma surgery datasets, demonstrating its broader diagnostic potential.

Contact: Michael Fujarski, M.Sc.