Segmentation and analysis of notochord cells in 3D microscope data

Boguslaw Obara*, Michael Veeman**, William Smithy**, B.S. Manjunath*
*Center for BioImage Informatics, Department Electrical and Computer Engineering, E-mail: {obara,manj} [at]
**Department of Molecular, Cellular and Developmental Biology, E-mail: {veeman,w smith} [at]
University of California, Santa Barbara, USA


The notochord is an embryonic structure that undergoes dramatic cell shape changes as it itself changes shape from a roughly isodiametric primordium of mesenchymal cells to a long extended rod. Ascidians (sea squirts) have small, compact embryos with a notochord consisting of only 40 cells, allowing comprehensive 3D confocal imaging that would be challenging in other chordate model organisms (Fig. 1(a)). Here we show a semiautomated method for the 3D segmentation of individual notochord cells, and present our prelimary findings applying this method to a multi-gigabyte confocal image collection of phalloidin-stained ascidian embryos. The proposed approach is based on mathematical morphology. The algorithm is able to reconstruct topological images and provide precise statistical data about the notochord cells (Fig. 1(b)). The proposed approach can be easily extended to any other tissues stained in a similar way.
[PDF] [BibTex]
Boguslaw Obara, Michael Veeman, William Smith and B.S. Manjunath,
Workshop on Bio-Image Informatics: Biological Imaging, Computer Vision and Data Mining, Santa Barbara, CA, USA, Jan. 2008.
Node ID: 499 , DB ID: 306 , Lab: VRL , Target: Workshop
Subject: [Detection on Images and Videos] « Look up more