Home // BIOTECHNO 2017, The Ninth International Conference on Bioinformatics, Biocomputational Systems and Biotechnologies // View article

Overlapping Cell Segmentation with Shape Prior

Authors:
Evan Yi-Wen Wang
Albert Kai-Sun Wong

Keywords: Cervical cells; Shape prior; Low-rank optimization.

Abstract:
Cell segmentation is a critical task in fully automatic computer cytology diagnosis. Overlapping cells pose a major challenge to cell segmentation because of blurred edges and inhomogeneous cytoplasm. In this paper, we propose a novel energy functional and formulate the boundary identification problem as a problem to minimize this energy functional together with a shape prior. The terms of our energy functional are designed to exploit both the edge and regional features, and the shape prior is derived from low rank optimization. For the non-convex minimization problem, we formulate an efficient monotone Accelerated Proximal Gradient with Linesearch (APGL) algorithm. We evaluate our proposed method using the International Symposium on Biomedical Imaging (ISBI) 2014 and 2015 challenge datasets. Results demonstrate that our method produces competitive accuracy compared to state-of-theart methods.

Pages: 10 to 15

Copyright: Copyright (c) IARIA, 2017

Publication date: May 21, 2017

Published in: conference

ISSN: 2308-4383

ISBN: 978-1-61208-560-9

Location: Barcelona, Spain

Dates: from May 21, 2017 to May 25, 2017