Classification of prostate histopathology images based on multifractal analysis

Chamidu Atupelage, Hiroshi Nagahashi, Masahiro Yamaguchi, Tokiya Abe, Akinori Hashiguchi, Michiie Sakamoto

研究成果: Article

3 引用 (Scopus)

抄録

Histopathology is a microscopic anatomical study of body tissues and widely used as a cancer diagnosing method. Generally, pathologists examine the structural deviation of cellular and sub-cellular components to diagnose the malignancy of body tissues. These judgments may often subjective to pathologists' skills and personal experiences. However, computational diagnosis tools may circumvent these limitations and improve the reliability of the diagnosis decisions. This paper proposes a prostate image classification method by extracting textural behavior using multifractal analysis. Fractal geometry is used to describe the complexity of self-similar structures as a non-integer exponent called fractal dimension. Natural complex structures (or images) are not self-similar, thus a single exponent (the fractal dimension) may not be adequate to describe the complexity of such structures. Multifractal analysis technique has been introduced to describe the complexity as a spectrum of fractal dimensions. Based on multifractal computation of digital imaging, we obtain two textural feature descriptors; i) local irregularity: α and ii) global regularity: f (α). We exploit these multifractal feature descriptors with a texton dictionary based classification model to discriminate cancer/non-cancer tissues of histopathology images of H&E stained prostate biopsy specimens. Moreover, we examine other three feature descriptors; Gabor filter bank, LM filter bank and Haralick features to benchmark the performance of the proposed method. Experiment results indicated that the performance of the proposed multifractal feature descriptor outperforms the other feature descriptors by achieving over 94% of correct classification accuracy.

元の言語English
ページ(範囲)3037-3045
ページ数9
ジャーナルIEICE Transactions on Information and Systems
E95-D
発行部数12
DOI
出版物ステータスPublished - 2012 12

Fingerprint

Fractal dimension
Filter banks
Tissue
Gabor filters
Biopsy
Image classification
Glossaries
Fractals
Imaging techniques
Geometry
Experiments

ASJC Scopus subject areas

  • Electrical and Electronic Engineering
  • Software
  • Artificial Intelligence
  • Hardware and Architecture
  • Computer Vision and Pattern Recognition

これを引用

Classification of prostate histopathology images based on multifractal analysis. / Atupelage, Chamidu; Nagahashi, Hiroshi; Yamaguchi, Masahiro; Abe, Tokiya; Hashiguchi, Akinori; Sakamoto, Michiie.

:: IEICE Transactions on Information and Systems, 巻 E95-D, 番号 12, 12.2012, p. 3037-3045.

研究成果: Article

Atupelage, Chamidu ; Nagahashi, Hiroshi ; Yamaguchi, Masahiro ; Abe, Tokiya ; Hashiguchi, Akinori ; Sakamoto, Michiie. / Classification of prostate histopathology images based on multifractal analysis. :: IEICE Transactions on Information and Systems. 2012 ; 巻 E95-D, 番号 12. pp. 3037-3045.
@article{072ce598b985446ca5c219bb718d5f89,
title = "Classification of prostate histopathology images based on multifractal analysis",
abstract = "Histopathology is a microscopic anatomical study of body tissues and widely used as a cancer diagnosing method. Generally, pathologists examine the structural deviation of cellular and sub-cellular components to diagnose the malignancy of body tissues. These judgments may often subjective to pathologists' skills and personal experiences. However, computational diagnosis tools may circumvent these limitations and improve the reliability of the diagnosis decisions. This paper proposes a prostate image classification method by extracting textural behavior using multifractal analysis. Fractal geometry is used to describe the complexity of self-similar structures as a non-integer exponent called fractal dimension. Natural complex structures (or images) are not self-similar, thus a single exponent (the fractal dimension) may not be adequate to describe the complexity of such structures. Multifractal analysis technique has been introduced to describe the complexity as a spectrum of fractal dimensions. Based on multifractal computation of digital imaging, we obtain two textural feature descriptors; i) local irregularity: α and ii) global regularity: f (α). We exploit these multifractal feature descriptors with a texton dictionary based classification model to discriminate cancer/non-cancer tissues of histopathology images of H&E stained prostate biopsy specimens. Moreover, we examine other three feature descriptors; Gabor filter bank, LM filter bank and Haralick features to benchmark the performance of the proposed method. Experiment results indicated that the performance of the proposed multifractal feature descriptor outperforms the other feature descriptors by achieving over 94{\%} of correct classification accuracy.",
keywords = "Classification, Feature descriptors, Fractal geometry, Histopathology, Multifractal, Prostate cancer",
author = "Chamidu Atupelage and Hiroshi Nagahashi and Masahiro Yamaguchi and Tokiya Abe and Akinori Hashiguchi and Michiie Sakamoto",
year = "2012",
month = "12",
doi = "10.1587/transinf.E95.D.3037",
language = "English",
volume = "E95-D",
pages = "3037--3045",
journal = "IEICE Transactions on Information and Systems",
issn = "0916-8532",
publisher = "Maruzen Co., Ltd/Maruzen Kabushikikaisha",
number = "12",

}

TY - JOUR

T1 - Classification of prostate histopathology images based on multifractal analysis

AU - Atupelage, Chamidu

AU - Nagahashi, Hiroshi

AU - Yamaguchi, Masahiro

AU - Abe, Tokiya

AU - Hashiguchi, Akinori

AU - Sakamoto, Michiie

PY - 2012/12

Y1 - 2012/12

N2 - Histopathology is a microscopic anatomical study of body tissues and widely used as a cancer diagnosing method. Generally, pathologists examine the structural deviation of cellular and sub-cellular components to diagnose the malignancy of body tissues. These judgments may often subjective to pathologists' skills and personal experiences. However, computational diagnosis tools may circumvent these limitations and improve the reliability of the diagnosis decisions. This paper proposes a prostate image classification method by extracting textural behavior using multifractal analysis. Fractal geometry is used to describe the complexity of self-similar structures as a non-integer exponent called fractal dimension. Natural complex structures (or images) are not self-similar, thus a single exponent (the fractal dimension) may not be adequate to describe the complexity of such structures. Multifractal analysis technique has been introduced to describe the complexity as a spectrum of fractal dimensions. Based on multifractal computation of digital imaging, we obtain two textural feature descriptors; i) local irregularity: α and ii) global regularity: f (α). We exploit these multifractal feature descriptors with a texton dictionary based classification model to discriminate cancer/non-cancer tissues of histopathology images of H&E stained prostate biopsy specimens. Moreover, we examine other three feature descriptors; Gabor filter bank, LM filter bank and Haralick features to benchmark the performance of the proposed method. Experiment results indicated that the performance of the proposed multifractal feature descriptor outperforms the other feature descriptors by achieving over 94% of correct classification accuracy.

AB - Histopathology is a microscopic anatomical study of body tissues and widely used as a cancer diagnosing method. Generally, pathologists examine the structural deviation of cellular and sub-cellular components to diagnose the malignancy of body tissues. These judgments may often subjective to pathologists' skills and personal experiences. However, computational diagnosis tools may circumvent these limitations and improve the reliability of the diagnosis decisions. This paper proposes a prostate image classification method by extracting textural behavior using multifractal analysis. Fractal geometry is used to describe the complexity of self-similar structures as a non-integer exponent called fractal dimension. Natural complex structures (or images) are not self-similar, thus a single exponent (the fractal dimension) may not be adequate to describe the complexity of such structures. Multifractal analysis technique has been introduced to describe the complexity as a spectrum of fractal dimensions. Based on multifractal computation of digital imaging, we obtain two textural feature descriptors; i) local irregularity: α and ii) global regularity: f (α). We exploit these multifractal feature descriptors with a texton dictionary based classification model to discriminate cancer/non-cancer tissues of histopathology images of H&E stained prostate biopsy specimens. Moreover, we examine other three feature descriptors; Gabor filter bank, LM filter bank and Haralick features to benchmark the performance of the proposed method. Experiment results indicated that the performance of the proposed multifractal feature descriptor outperforms the other feature descriptors by achieving over 94% of correct classification accuracy.

KW - Classification

KW - Feature descriptors

KW - Fractal geometry

KW - Histopathology

KW - Multifractal

KW - Prostate cancer

UR - http://www.scopus.com/inward/record.url?scp=84870692572&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84870692572&partnerID=8YFLogxK

U2 - 10.1587/transinf.E95.D.3037

DO - 10.1587/transinf.E95.D.3037

M3 - Article

AN - SCOPUS:84870692572

VL - E95-D

SP - 3037

EP - 3045

JO - IEICE Transactions on Information and Systems

JF - IEICE Transactions on Information and Systems

SN - 0916-8532

IS - 12

ER -