Tubes are more than infrastructure. The formation of ductal tubes in the embryonic pancreas occurs in a unique manner requiring extensive remodeling of a network structure into a hierarchical tree-like structure, thereby optimizing flow in the tubes. While it is well-known that tubes form crucial transportation systems, recent research on ductal tubes in the pancreas indicates that they also guide specification of cellular fate. We hypothesise that topological events during embryonic development, such as the generation and closing of loops in the tubular network, are related to the maturation of insulin-producing beta-cells residing in the tubes. We aim to detect and quantify the topological events that remodels the embryonic pancreas, and to relate them to the spatial distribution of developing beta-cells.
The detection and quantification of topological events relies on the development of image segmentation tools as well as highly scalable topological data analysis algorithms. Moreover, we will also develop novel algorithms for tracking topological features over time. As the large-scale quantification of topological features will be based on automatic (and hence, error-prone) image segmentations in high resolution time-lapse movies, we will model
their expected uncertainty based on manual annotations.
Once we have detected and quantified topological features and events, we will develop spatial statistics methods to answer the original biological question: How is the pancreatic ductal network remodeled? and are topological events in the developing tubular network correlated with the maturation of beta-cells?