Hierarchical, model-based merging of multiple fragments for improved three-dimensional segmentation of nuclei

Gang Lin, Monica K. Chawla, Kathy Olson, John F. Guzowski, Carol A Barnes, Badrinath Roysam

Research output: Contribution to journalArticle

94 Citations (Scopus)

Abstract

Background: Automated segmentation of fluorescently labeled cell nuclei in three-dimensional confocal images is essential for numerous studies, e.g., spatiotemporal fluorescence in situ hybridization quantification of immediate early gene transcription. High accuracy and automation levels are required in high-throughput and large-scale studies. Common sources of segmentation error include tight clustering and fragmentation of nuclei. Previous region-based methods are limited because they perform merging of two nuclear fragments at a time. To achieve higher accuracy without sacrificing scale, more sophisticated yet computationally efficient algorithms are needed. Methods: A recursive tree-based algorithm that can consider multiple object fragments simultaneously is described. Starting with oversegmented data, it searches efficiently for the optimal merging pattern guided by a quantitative scoring criterion based on object modeling. Computation is bounded by limiting the depth of the merging tree. Results: The proposed method was found to perform consistently better, achieving merging accuracy in the range of 92% to 100% compared with our previous algorithm, which varied in the range of 75% to 97%, even with a modest merging tree depth of 3. The overall average accuracy improved from 90% to 96%, with roughly the same computational cost for a set of representative images drawn from the CA1, CA3, and parietal cortex regions of the rat hippocampus. Conclusion: Hierarchical tree model-based algorithms significantly improve the accuracy of automated nuclear segmentation without sacrificing speed.

Original languageEnglish (US)
Pages (from-to)20-33
Number of pages14
JournalCytometry Part A
Volume63
Issue number1
DOIs
StatePublished - Jan 2005

Fingerprint

Parietal Lobe
Genetic Transcription
Immediate-Early Genes
Three-Dimensional Imaging
Automation
Cell Nucleus
Fluorescence In Situ Hybridization
Cluster Analysis
Hippocampus
Research Design
Costs and Cost Analysis

Keywords

  • Cell counting
  • Confocal microscopy
  • Fluorescence in situ hybridization quantification
  • Hierarchical
  • Image segmentation
  • Model based
  • Object features
  • Region merging
  • Three-dimensional image analysis
  • Watershed segmentation

ASJC Scopus subject areas

  • Hematology
  • Cell Biology
  • Pathology and Forensic Medicine
  • Biophysics
  • Endocrinology

Cite this

Hierarchical, model-based merging of multiple fragments for improved three-dimensional segmentation of nuclei. / Lin, Gang; Chawla, Monica K.; Olson, Kathy; Guzowski, John F.; Barnes, Carol A; Roysam, Badrinath.

In: Cytometry Part A, Vol. 63, No. 1, 01.2005, p. 20-33.

Research output: Contribution to journalArticle

Lin, Gang ; Chawla, Monica K. ; Olson, Kathy ; Guzowski, John F. ; Barnes, Carol A ; Roysam, Badrinath. / Hierarchical, model-based merging of multiple fragments for improved three-dimensional segmentation of nuclei. In: Cytometry Part A. 2005 ; Vol. 63, No. 1. pp. 20-33.
@article{74d513e84b1e4c3984cb81e690187433,
title = "Hierarchical, model-based merging of multiple fragments for improved three-dimensional segmentation of nuclei",
abstract = "Background: Automated segmentation of fluorescently labeled cell nuclei in three-dimensional confocal images is essential for numerous studies, e.g., spatiotemporal fluorescence in situ hybridization quantification of immediate early gene transcription. High accuracy and automation levels are required in high-throughput and large-scale studies. Common sources of segmentation error include tight clustering and fragmentation of nuclei. Previous region-based methods are limited because they perform merging of two nuclear fragments at a time. To achieve higher accuracy without sacrificing scale, more sophisticated yet computationally efficient algorithms are needed. Methods: A recursive tree-based algorithm that can consider multiple object fragments simultaneously is described. Starting with oversegmented data, it searches efficiently for the optimal merging pattern guided by a quantitative scoring criterion based on object modeling. Computation is bounded by limiting the depth of the merging tree. Results: The proposed method was found to perform consistently better, achieving merging accuracy in the range of 92{\%} to 100{\%} compared with our previous algorithm, which varied in the range of 75{\%} to 97{\%}, even with a modest merging tree depth of 3. The overall average accuracy improved from 90{\%} to 96{\%}, with roughly the same computational cost for a set of representative images drawn from the CA1, CA3, and parietal cortex regions of the rat hippocampus. Conclusion: Hierarchical tree model-based algorithms significantly improve the accuracy of automated nuclear segmentation without sacrificing speed.",
keywords = "Cell counting, Confocal microscopy, Fluorescence in situ hybridization quantification, Hierarchical, Image segmentation, Model based, Object features, Region merging, Three-dimensional image analysis, Watershed segmentation",
author = "Gang Lin and Chawla, {Monica K.} and Kathy Olson and Guzowski, {John F.} and Barnes, {Carol A} and Badrinath Roysam",
year = "2005",
month = "1",
doi = "10.1002/cyto.a.20099",
language = "English (US)",
volume = "63",
pages = "20--33",
journal = "Cytometry. Part A : the journal of the International Society for Analytical Cytology",
issn = "1552-4922",
publisher = "Wiley-Liss Inc.",
number = "1",

}

TY - JOUR

T1 - Hierarchical, model-based merging of multiple fragments for improved three-dimensional segmentation of nuclei

AU - Lin, Gang

AU - Chawla, Monica K.

AU - Olson, Kathy

AU - Guzowski, John F.

AU - Barnes, Carol A

AU - Roysam, Badrinath

PY - 2005/1

Y1 - 2005/1

N2 - Background: Automated segmentation of fluorescently labeled cell nuclei in three-dimensional confocal images is essential for numerous studies, e.g., spatiotemporal fluorescence in situ hybridization quantification of immediate early gene transcription. High accuracy and automation levels are required in high-throughput and large-scale studies. Common sources of segmentation error include tight clustering and fragmentation of nuclei. Previous region-based methods are limited because they perform merging of two nuclear fragments at a time. To achieve higher accuracy without sacrificing scale, more sophisticated yet computationally efficient algorithms are needed. Methods: A recursive tree-based algorithm that can consider multiple object fragments simultaneously is described. Starting with oversegmented data, it searches efficiently for the optimal merging pattern guided by a quantitative scoring criterion based on object modeling. Computation is bounded by limiting the depth of the merging tree. Results: The proposed method was found to perform consistently better, achieving merging accuracy in the range of 92% to 100% compared with our previous algorithm, which varied in the range of 75% to 97%, even with a modest merging tree depth of 3. The overall average accuracy improved from 90% to 96%, with roughly the same computational cost for a set of representative images drawn from the CA1, CA3, and parietal cortex regions of the rat hippocampus. Conclusion: Hierarchical tree model-based algorithms significantly improve the accuracy of automated nuclear segmentation without sacrificing speed.

AB - Background: Automated segmentation of fluorescently labeled cell nuclei in three-dimensional confocal images is essential for numerous studies, e.g., spatiotemporal fluorescence in situ hybridization quantification of immediate early gene transcription. High accuracy and automation levels are required in high-throughput and large-scale studies. Common sources of segmentation error include tight clustering and fragmentation of nuclei. Previous region-based methods are limited because they perform merging of two nuclear fragments at a time. To achieve higher accuracy without sacrificing scale, more sophisticated yet computationally efficient algorithms are needed. Methods: A recursive tree-based algorithm that can consider multiple object fragments simultaneously is described. Starting with oversegmented data, it searches efficiently for the optimal merging pattern guided by a quantitative scoring criterion based on object modeling. Computation is bounded by limiting the depth of the merging tree. Results: The proposed method was found to perform consistently better, achieving merging accuracy in the range of 92% to 100% compared with our previous algorithm, which varied in the range of 75% to 97%, even with a modest merging tree depth of 3. The overall average accuracy improved from 90% to 96%, with roughly the same computational cost for a set of representative images drawn from the CA1, CA3, and parietal cortex regions of the rat hippocampus. Conclusion: Hierarchical tree model-based algorithms significantly improve the accuracy of automated nuclear segmentation without sacrificing speed.

KW - Cell counting

KW - Confocal microscopy

KW - Fluorescence in situ hybridization quantification

KW - Hierarchical

KW - Image segmentation

KW - Model based

KW - Object features

KW - Region merging

KW - Three-dimensional image analysis

KW - Watershed segmentation

UR - http://www.scopus.com/inward/record.url?scp=13244256957&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=13244256957&partnerID=8YFLogxK

U2 - 10.1002/cyto.a.20099

DO - 10.1002/cyto.a.20099

M3 - Article

C2 - 15584021

AN - SCOPUS:13244256957

VL - 63

SP - 20

EP - 33

JO - Cytometry. Part A : the journal of the International Society for Analytical Cytology

JF - Cytometry. Part A : the journal of the International Society for Analytical Cytology

SN - 1552-4922

IS - 1

ER -