EPFL
 Biomedical Imaging GroupSTI
EPFL
  Publications
English only   BIG > Publications > Cell Segmentation


 CONTENTS
 Home Page
 News & Events
 People
 Publications
 Tutorials and Reviews
 Research
 Demos
 Download Algorithms

 DOWNLOAD
 PDF
 Postscript
 All BibTeX References

Segmentation of Adjacent Cells in a C. elegans Embryo Using a Multi-Compartment Active Contour

A. Badoual, A. Galan, D. Sage, M. Unser

Proceedings of the 2019 Quantitative Bioimaging Conference (QBI'19), Rennes, French Republic, January 9-11, 2019, pp. 68.



Active contours—a.k.a snakes—are popular models for the segmentation of biological structures. They consist in an initial closed curve that evolves towards the boundary of the object of interest. The evolution is driven by the minimization of an application-dependent energy term.

We present a framework to automatically segment a compact cluster of cells in a C. elegans embryo in 2D light sheet fluorescence microscopy images. For such a purpose, we designed a multi-compartment active contour to fit a structure composed of several adjacent cell membranes. We take advantage of subdivision schemes: for each compartment we recursively apply a refinement process to an initial set of few points to produce a continuous limit curve. The final multi-compartment curve is then deformed in a global manner using a suitable ridge-based energy attracted by the fluorescent membranes. We automatize the initialization of the snake by constructing a Voronoi diagram around the labeled nuclei. This framework has several advantages: 1) the coherent structure of the multi-compartment curve is maintained even on membranes with low or missing fluorescent information. It avoids the leaking problem which afflicts classical image-processing methods; 2) it requires few parameters (i.e., control points), which results in faster optimization and better robustness; 3) the user can intuitively interact with the curve by modifying some control points.

We validated our framework both on synthetic and real fluorescence microscopy images showing that our algorithm is robust to dim staining and to high levels of noise.

We implemented this framework as a user-friendly plugin for the bioimaging software package Icy.


@INPROCEEDINGS(http://bigwww.epfl.ch/publications/badoual1901.html,
AUTHOR="Badoual, A. and Galan, A. and Sage, D. and Unser, M.",
TITLE="Segmentation of Adjacent Cells in a {\textit{C. elegans}} Embryo
        Using a Multi-Compartment Active Contour",
BOOKTITLE="2019 Quantitative {B}io{I}maging Conference ({QBIC'19})",
YEAR="2019",
editor="",
volume="",
series="",
pages="68",
address="Rennes, French Republic",
month="January 9-11,",
organization="",
publisher="",
note="paper no. 40")

© 2019 ANR. Personal use of this material is permitted. However, permission to reprint/republish this material for advertising or promotional purposes or for creating new collective works for resale or redistribution to servers or lists, or to reuse any copyrighted component of this work in other works must be obtained from ANR.
This material is presented to ensure timely dissemination of scholarly and technical work. Copyright and all rights therein are retained by authors or by other copyright holders. All persons copying this information are expected to adhere to the terms and constraints invoked by each author's copyright. In most cases, these works may not be reposted without the explicit permission of the copyright holder.