Learning the topological properties of brain tumors

Cigdem Demir, Sakir Gultekin, Bülent Yener

Research output: Contribution to journalArticle

39 Citations (Scopus)

Abstract

This work presents a graph-based representation (a.k.a., cell-graph) of histopathological images for automated cancer diagnosis by probabilistically assigning a link between a pair of cells (or cell clusters). Since the node set of a cell-graph can include a cluster of cells as well as individual ones, it enables working with low-cost, low-magnification photomicrographs. The contributions of this work are twofold. First, it is shown that without establishing a pairwise spatial relation between the cells (i.e., the edges of a cellgraph), neither the spatial distribution of the cells nor the texture analysis of the images yields accurate results for tissue level diagnosis of brain cancer called malignant glioma. Second, this work defines a set of global metrics by processing the entire cell-graph to capture tissue level information coded into the histopathological images. In this work, the results are obtained on the photomicrographs of 646 archival brain biopsy samples of 60 different patients. It is shown that the global metrics of cell-graphs distinguish cancerous tissues from noncancerous ones with high accuracy (at least 99 percent accuracy for healthy tissues with lower cellular density level, and at least 92 percent accuracy for benign tissues with similar high cellular density level such as nonneoplastic reactive/inflammatory conditions).

Original languageEnglish (US)
Pages (from-to)262-269
Number of pages8
JournalIEEE/ACM Transactions on Computational Biology and Bioinformatics
Volume2
Issue number3
DOIs
StatePublished - Jul 2005

Fingerprint

Brain Tumor
Topological Properties
Brain Neoplasms
Tumors
Brain
learning
Learning
Tissue
brain
neoplasms
Cell
cells
Graph in graph theory
Biopsy
Percent
Cancer
Spatial distribution
Textures
Cells
Metric

Keywords

  • Graph theory
  • Image representation
  • Machine learning
  • Medical information systems
  • Model development

ASJC Scopus subject areas

  • Engineering(all)
  • Agricultural and Biological Sciences (miscellaneous)

Cite this

Learning the topological properties of brain tumors. / Demir, Cigdem; Gultekin, Sakir; Yener, Bülent.

In: IEEE/ACM Transactions on Computational Biology and Bioinformatics, Vol. 2, No. 3, 07.2005, p. 262-269.

Research output: Contribution to journalArticle

Demir, Cigdem ; Gultekin, Sakir ; Yener, Bülent. / Learning the topological properties of brain tumors. In: IEEE/ACM Transactions on Computational Biology and Bioinformatics. 2005 ; Vol. 2, No. 3. pp. 262-269.
@article{0ad177cb2d00493b8f49666c627f2767,
title = "Learning the topological properties of brain tumors",
abstract = "This work presents a graph-based representation (a.k.a., cell-graph) of histopathological images for automated cancer diagnosis by probabilistically assigning a link between a pair of cells (or cell clusters). Since the node set of a cell-graph can include a cluster of cells as well as individual ones, it enables working with low-cost, low-magnification photomicrographs. The contributions of this work are twofold. First, it is shown that without establishing a pairwise spatial relation between the cells (i.e., the edges of a cellgraph), neither the spatial distribution of the cells nor the texture analysis of the images yields accurate results for tissue level diagnosis of brain cancer called malignant glioma. Second, this work defines a set of global metrics by processing the entire cell-graph to capture tissue level information coded into the histopathological images. In this work, the results are obtained on the photomicrographs of 646 archival brain biopsy samples of 60 different patients. It is shown that the global metrics of cell-graphs distinguish cancerous tissues from noncancerous ones with high accuracy (at least 99 percent accuracy for healthy tissues with lower cellular density level, and at least 92 percent accuracy for benign tissues with similar high cellular density level such as nonneoplastic reactive/inflammatory conditions).",
keywords = "Graph theory, Image representation, Machine learning, Medical information systems, Model development",
author = "Cigdem Demir and Sakir Gultekin and B{\"u}lent Yener",
year = "2005",
month = "7",
doi = "10.1109/TCBB.2005.42",
language = "English (US)",
volume = "2",
pages = "262--269",
journal = "IEEE/ACM Transactions on Computational Biology and Bioinformatics",
issn = "1545-5963",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
number = "3",

}

TY - JOUR

T1 - Learning the topological properties of brain tumors

AU - Demir, Cigdem

AU - Gultekin, Sakir

AU - Yener, Bülent

PY - 2005/7

Y1 - 2005/7

N2 - This work presents a graph-based representation (a.k.a., cell-graph) of histopathological images for automated cancer diagnosis by probabilistically assigning a link between a pair of cells (or cell clusters). Since the node set of a cell-graph can include a cluster of cells as well as individual ones, it enables working with low-cost, low-magnification photomicrographs. The contributions of this work are twofold. First, it is shown that without establishing a pairwise spatial relation between the cells (i.e., the edges of a cellgraph), neither the spatial distribution of the cells nor the texture analysis of the images yields accurate results for tissue level diagnosis of brain cancer called malignant glioma. Second, this work defines a set of global metrics by processing the entire cell-graph to capture tissue level information coded into the histopathological images. In this work, the results are obtained on the photomicrographs of 646 archival brain biopsy samples of 60 different patients. It is shown that the global metrics of cell-graphs distinguish cancerous tissues from noncancerous ones with high accuracy (at least 99 percent accuracy for healthy tissues with lower cellular density level, and at least 92 percent accuracy for benign tissues with similar high cellular density level such as nonneoplastic reactive/inflammatory conditions).

AB - This work presents a graph-based representation (a.k.a., cell-graph) of histopathological images for automated cancer diagnosis by probabilistically assigning a link between a pair of cells (or cell clusters). Since the node set of a cell-graph can include a cluster of cells as well as individual ones, it enables working with low-cost, low-magnification photomicrographs. The contributions of this work are twofold. First, it is shown that without establishing a pairwise spatial relation between the cells (i.e., the edges of a cellgraph), neither the spatial distribution of the cells nor the texture analysis of the images yields accurate results for tissue level diagnosis of brain cancer called malignant glioma. Second, this work defines a set of global metrics by processing the entire cell-graph to capture tissue level information coded into the histopathological images. In this work, the results are obtained on the photomicrographs of 646 archival brain biopsy samples of 60 different patients. It is shown that the global metrics of cell-graphs distinguish cancerous tissues from noncancerous ones with high accuracy (at least 99 percent accuracy for healthy tissues with lower cellular density level, and at least 92 percent accuracy for benign tissues with similar high cellular density level such as nonneoplastic reactive/inflammatory conditions).

KW - Graph theory

KW - Image representation

KW - Machine learning

KW - Medical information systems

KW - Model development

UR - http://www.scopus.com/inward/record.url?scp=27544512013&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=27544512013&partnerID=8YFLogxK

U2 - 10.1109/TCBB.2005.42

DO - 10.1109/TCBB.2005.42

M3 - Article

C2 - 17044189

AN - SCOPUS:27544512013

VL - 2

SP - 262

EP - 269

JO - IEEE/ACM Transactions on Computational Biology and Bioinformatics

JF - IEEE/ACM Transactions on Computational Biology and Bioinformatics

SN - 1545-5963

IS - 3

ER -