Scalable Robust Graph and Feature Extraction for Arbitrary Vessel Networks in Large Volumetric Datasets
Authors:
Dominik Drees,
Aaron Scherzinger,
René Hägerling,
Friedemann Kiefer,
Xiaoyi Jiang
Abstract:
Recent advances in 3D imaging technologies provide novel insights to researchers and reveal finer and more detail of examined specimen, especially in the biomedical domain, but also impose huge challenges regarding scalability for automated analysis algorithms due to rapidly increasing dataset sizes. In particular, existing research towards automated vessel network analysis does not consider memor…
▽ More
Recent advances in 3D imaging technologies provide novel insights to researchers and reveal finer and more detail of examined specimen, especially in the biomedical domain, but also impose huge challenges regarding scalability for automated analysis algorithms due to rapidly increasing dataset sizes. In particular, existing research towards automated vessel network analysis does not consider memory requirements of proposed algorithms and often generates a large number of spurious branches for structures consisting of many voxels. Additionally, very often these algorithms have further restrictions such as the limitation to tree topologies or relying on the properties of specific image modalities. We present a scalable pipeline (in terms of computational cost, required main memory and robustness) that extracts an annotated abstract graph representation from the foreground segmentation of vessel networks of arbitrary topology and vessel shape. Only a single, dimensionless, a-priori determinable parameter is required. By careful engineering of individual pipeline stages and a novel iterative refinement scheme we are, for the first time, able to analyze the topology of volumes of roughly 1TB on commodity hardware. An implementation of the presented pipeline is publicly available in version 5.1 of the volume rendering and processing engine Voreen (https://www.uni-muenster.de/Voreen/).
△ Less
Submitted 5 February, 2021;
originally announced February 2021.
