Cell Detection From Redundant Candidate Regions Under Nonoverlapping Constraints

Ryoma Bise, Yoichi Sato

Research output: Contribution to journalArticle

13 Citations (Scopus)

Abstract

Cell detection in microscopy images is essential for automated cell behavior analysis including cell shape analysis and cell tracking. Robust cell detection in high-density and low-contrast images is still challenging since cells often touch and partially overlap, forming a cell cluster with blurry intercellular boundaries. In such cases, current methods tend to detect multiple cells as a cluster. If the control parameters are adjusted to separate the touching cells, other problems often occur: a single cell may be segmented into several regions, and cells in low-intensity regions may not be detected. To solve these problems, we first detect redundant candidate regions, which include many false positives but in turn very few false negatives, by allowing candidate regions to overlap with each other. Next, the score for how likely the candidate region contains the main part of a single cell is computed for each cell candidate using supervised learning. Then we select an optimal set of cell regions from the redundant regions under nonoverlapping constraints, where each selected region looks like a single cell and the selected regions do not overlap. We formulate this problem of optimal region selection as a binary linear programming problem under nonoverlapping constraints. We demonstrated the effectiveness of our method for several types of cells in microscopy images. Our method performed better than five representative methods, achieving an F-measure of over 0.9 for all data sets. Experimental application of the proposed method to 3-D images demonstrated that also works well for 3-D cell detection.

Original languageEnglish
Article number7006809
Pages (from-to)1417-1427
Number of pages11
JournalIEEE Transactions on Medical Imaging
Volume34
Issue number7
DOIs
Publication statusPublished - Jul 1 2015
Externally publishedYes

Fingerprint

Microscopic examination
Supervised learning
Linear programming
Microscopy
Cell Tracking
Linear Programming
Somatostatin-Secreting Cells
Three-Dimensional Imaging
Cell Shape
Touch
Learning

All Science Journal Classification (ASJC) codes

  • Software
  • Radiological and Ultrasound Technology
  • Computer Science Applications
  • Electrical and Electronic Engineering

Cite this

Cell Detection From Redundant Candidate Regions Under Nonoverlapping Constraints. / Bise, Ryoma; Sato, Yoichi.

In: IEEE Transactions on Medical Imaging, Vol. 34, No. 7, 7006809, 01.07.2015, p. 1417-1427.

Research output: Contribution to journalArticle

@article{dc0657a1760442ef91314f2127e2969c,
title = "Cell Detection From Redundant Candidate Regions Under Nonoverlapping Constraints",
abstract = "Cell detection in microscopy images is essential for automated cell behavior analysis including cell shape analysis and cell tracking. Robust cell detection in high-density and low-contrast images is still challenging since cells often touch and partially overlap, forming a cell cluster with blurry intercellular boundaries. In such cases, current methods tend to detect multiple cells as a cluster. If the control parameters are adjusted to separate the touching cells, other problems often occur: a single cell may be segmented into several regions, and cells in low-intensity regions may not be detected. To solve these problems, we first detect redundant candidate regions, which include many false positives but in turn very few false negatives, by allowing candidate regions to overlap with each other. Next, the score for how likely the candidate region contains the main part of a single cell is computed for each cell candidate using supervised learning. Then we select an optimal set of cell regions from the redundant regions under nonoverlapping constraints, where each selected region looks like a single cell and the selected regions do not overlap. We formulate this problem of optimal region selection as a binary linear programming problem under nonoverlapping constraints. We demonstrated the effectiveness of our method for several types of cells in microscopy images. Our method performed better than five representative methods, achieving an F-measure of over 0.9 for all data sets. Experimental application of the proposed method to 3-D images demonstrated that also works well for 3-D cell detection.",
author = "Ryoma Bise and Yoichi Sato",
year = "2015",
month = "7",
day = "1",
doi = "10.1109/TMI.2015.2391095",
language = "English",
volume = "34",
pages = "1417--1427",
journal = "IEEE Transactions on Medical Imaging",
issn = "0278-0062",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
number = "7",

}

TY - JOUR

T1 - Cell Detection From Redundant Candidate Regions Under Nonoverlapping Constraints

AU - Bise, Ryoma

AU - Sato, Yoichi

PY - 2015/7/1

Y1 - 2015/7/1

N2 - Cell detection in microscopy images is essential for automated cell behavior analysis including cell shape analysis and cell tracking. Robust cell detection in high-density and low-contrast images is still challenging since cells often touch and partially overlap, forming a cell cluster with blurry intercellular boundaries. In such cases, current methods tend to detect multiple cells as a cluster. If the control parameters are adjusted to separate the touching cells, other problems often occur: a single cell may be segmented into several regions, and cells in low-intensity regions may not be detected. To solve these problems, we first detect redundant candidate regions, which include many false positives but in turn very few false negatives, by allowing candidate regions to overlap with each other. Next, the score for how likely the candidate region contains the main part of a single cell is computed for each cell candidate using supervised learning. Then we select an optimal set of cell regions from the redundant regions under nonoverlapping constraints, where each selected region looks like a single cell and the selected regions do not overlap. We formulate this problem of optimal region selection as a binary linear programming problem under nonoverlapping constraints. We demonstrated the effectiveness of our method for several types of cells in microscopy images. Our method performed better than five representative methods, achieving an F-measure of over 0.9 for all data sets. Experimental application of the proposed method to 3-D images demonstrated that also works well for 3-D cell detection.

AB - Cell detection in microscopy images is essential for automated cell behavior analysis including cell shape analysis and cell tracking. Robust cell detection in high-density and low-contrast images is still challenging since cells often touch and partially overlap, forming a cell cluster with blurry intercellular boundaries. In such cases, current methods tend to detect multiple cells as a cluster. If the control parameters are adjusted to separate the touching cells, other problems often occur: a single cell may be segmented into several regions, and cells in low-intensity regions may not be detected. To solve these problems, we first detect redundant candidate regions, which include many false positives but in turn very few false negatives, by allowing candidate regions to overlap with each other. Next, the score for how likely the candidate region contains the main part of a single cell is computed for each cell candidate using supervised learning. Then we select an optimal set of cell regions from the redundant regions under nonoverlapping constraints, where each selected region looks like a single cell and the selected regions do not overlap. We formulate this problem of optimal region selection as a binary linear programming problem under nonoverlapping constraints. We demonstrated the effectiveness of our method for several types of cells in microscopy images. Our method performed better than five representative methods, achieving an F-measure of over 0.9 for all data sets. Experimental application of the proposed method to 3-D images demonstrated that also works well for 3-D cell detection.

UR - http://www.scopus.com/inward/record.url?scp=84927652662&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84927652662&partnerID=8YFLogxK

U2 - 10.1109/TMI.2015.2391095

DO - 10.1109/TMI.2015.2391095

M3 - Article

AN - SCOPUS:84927652662

VL - 34

SP - 1417

EP - 1427

JO - IEEE Transactions on Medical Imaging

JF - IEEE Transactions on Medical Imaging

SN - 0278-0062

IS - 7

M1 - 7006809

ER -