Format

Send to

Choose Destination
See comment in PubMed Commons below
Conf Proc IEEE Eng Med Biol Soc. 2013;2013:6087-90. doi: 10.1109/EMBC.2013.6610941.

Cell nuclei segmentation in fluorescence microscopy images using inter- and intra-region discriminative information.

Abstract

Automated segmentation of cell nuclei in microscopic images is critical to high throughput analysis of the ever increasing amount of data. Although cell nuclei are generally visually distinguishable for human, automated segmentation faces challenges when there is significant intensity inhomogeneity among cell nuclei or in the background. In this paper, we propose an effective method for automated cell nucleus segmentation using a three-step approach. It first obtains an initial segmentation by extracting salient regions in the image, then reduces false positives using inter-region feature discrimination, and finally refines the boundary of the cell nuclei using intra-region contrast information. This method has been evaluated on two publicly available datasets of fluorescence microscopic images with 4009 cells, and has achieved superior performance compared to popular state of the art methods using established metrics.

PMID:
24111128
DOI:
10.1109/EMBC.2013.6610941
[Indexed for MEDLINE]
PubMed Commons home

PubMed Commons

0 comments
How to join PubMed Commons

    Supplemental Content

    Full text links

    Icon for IEEE Engineering in Medicine and Biology Society
    Loading ...
    Support Center