Premium
REAVER: An Improved Image Analysis Pipeline for Quantifying Microvascular Networks
Author(s) -
Corliss Bruce A.,
Doty Richard,
Matthews Corbin,
Peirce Shayn M.
Publication year - 2019
Publication title -
the faseb journal
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.709
H-Index - 277
eISSN - 1530-6860
pISSN - 0892-6638
DOI - 10.1096/fasebj.2019.33.1_supplement.525.7
Subject(s) - ground truth , pipeline (software) , segmentation , computer science , benchmark (surveying) , projection (relational algebra) , artificial intelligence , computer vision , cartography , algorithm , programming language , geography
Alterations in microvessel networks, including angiogenesis and capillary regression, play key roles in disease, injury, and development. Imaging of microvascular networks can capture their spatial structures, but effective study of network architecture requires methods to accurately quantify them. We present REAVER (Rapid Editable Analysis of Vessel Elements Routine), a tool to analyze blood vessel networks, and demonstrate that REAVER has superior performance compared to the state‐of‐the‐art1–3. Material and Methods Retinas from C57Bl/6J mice were labeled with IB4 Lectin Alexa Flour 647 and imaged with a Nikon TE‐2000E point scanning confocal microscope. A total of 36 z‐stacks from a variety of tissues were flattened to 2D images with a maximum intensity projection and used as a benchmark dataset. To establish ground truth, all images were manually analyzed in ImageJ. REAVER is written in MATLAB 2018a (Figure 1A). Results and Discussion REAVER demonstrated markedly higher performance in segmentation and quantification of vessel networks compared to currently available software packages. For vessel length, REAVER had lower mean relative error to ground truth (−1.4% ± 0.091) compared to Angioquant (62.8% ± 0.10), Angiotool (27.0% ± 0.08), and RAVE (43.1% ± 0.17) (Figure 1B). REAVER also had lower error to ground truth with branchpoint count (10.1% ± 0.3) compared to Angioquant (−649% ± 5.06), Angiotool (−341.2% ± 2.90), and RAVE (−1884% ± 12.36) (Figure 1C). Higher accuracy in quantification of vessel networks will allow for improved ability to discern alterations in vessel architecture for microvascular research. Support or Funding Information Funding: This work was supported by NIH R21 EY028868‐01 and The Hartwell Foundation (to S.M.P.). This abstract is from the Experimental Biology 2019 Meeting. There is no full text article associated with this abstract published in The FASEB Journal .