Histology image classification using supervised classification and multimodal fusion

Tao Meng, Lin Lin, Mei-Ling Shyu, Shu Ching Chen

Research output: Chapter in Book/Report/Conference proceedingConference contribution

27 Citations (Scopus)

Abstract

The fast development of microscopy imaging techniques nowadays promotes the generation of a large amount of data. These data are very crucial not only for theoretical biomedical research but also for clinical usage. In order to decrease the inter-intra observer variability and save the human effort on labeling and classifying these images, a lot of research efforts have been devoted to the development of algorithms for biomedical images. Among such efforts, histology image classification is one of the most important areas due to its broad applications in pathological diagnosis such as cancer diagnosis. To improve classification accuracy, most of the previous work focuses on extracting more features and building algorithms for a specific task. This paper proposes a framework based on the novel and robust Collateral Representative Subspace Projection Modeling (C-RSPM) supervised classification model for general histology image classification. In the proposed framework, a cell image is first divided into 25 blocks to reduce the spatial complexity of computation, and one C-RSPM model is built on each block set which contains blocks in the same location from different images. For each testing image, our proposed framework first classifies each of its blocks using the C-RSPM classification model built for that block set, and then applies a multimodal late fusion algorithm with a weighted majority voting strategy to decide the final class label of the whole image. Experimenting using three-fold cross validation with three benchmark histology data sets shows that the proposed framework outperforms other well-known classifiers in the comparison and gives better results than the highest accuracy reported previously.

Original languageEnglish
Title of host publicationProceedings - 2010 IEEE International Symposium on Multimedia, ISM 2010
Pages145-152
Number of pages8
DOIs
StatePublished - Dec 1 2010
Event2010 IEEE International Symposium on Multimedia, ISM 2010 - Taichung, Taiwan, Province of China
Duration: Dec 13 2010Dec 15 2010

Other

Other2010 IEEE International Symposium on Multimedia, ISM 2010
CountryTaiwan, Province of China
CityTaichung
Period12/13/1012/15/10

Fingerprint

Histology
Image classification
Fusion reactions
Labeling
Labels
Microscopic examination
Classifiers
Imaging techniques
Testing

Keywords

  • C-RSPM
  • Histology image classification
  • Multimodal fusion
  • Weighted majority voting algorithm

ASJC Scopus subject areas

  • Computer Graphics and Computer-Aided Design
  • Human-Computer Interaction

Cite this

Meng, T., Lin, L., Shyu, M-L., & Chen, S. C. (2010). Histology image classification using supervised classification and multimodal fusion. In Proceedings - 2010 IEEE International Symposium on Multimedia, ISM 2010 (pp. 145-152). [5693834] https://doi.org/10.1109/ISM.2010.29

Histology image classification using supervised classification and multimodal fusion. / Meng, Tao; Lin, Lin; Shyu, Mei-Ling; Chen, Shu Ching.

Proceedings - 2010 IEEE International Symposium on Multimedia, ISM 2010. 2010. p. 145-152 5693834.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Meng, T, Lin, L, Shyu, M-L & Chen, SC 2010, Histology image classification using supervised classification and multimodal fusion. in Proceedings - 2010 IEEE International Symposium on Multimedia, ISM 2010., 5693834, pp. 145-152, 2010 IEEE International Symposium on Multimedia, ISM 2010, Taichung, Taiwan, Province of China, 12/13/10. https://doi.org/10.1109/ISM.2010.29
Meng T, Lin L, Shyu M-L, Chen SC. Histology image classification using supervised classification and multimodal fusion. In Proceedings - 2010 IEEE International Symposium on Multimedia, ISM 2010. 2010. p. 145-152. 5693834 https://doi.org/10.1109/ISM.2010.29
Meng, Tao ; Lin, Lin ; Shyu, Mei-Ling ; Chen, Shu Ching. / Histology image classification using supervised classification and multimodal fusion. Proceedings - 2010 IEEE International Symposium on Multimedia, ISM 2010. 2010. pp. 145-152
@inproceedings{545a21b0cd8240d89d19fc43e30d1451,
title = "Histology image classification using supervised classification and multimodal fusion",
abstract = "The fast development of microscopy imaging techniques nowadays promotes the generation of a large amount of data. These data are very crucial not only for theoretical biomedical research but also for clinical usage. In order to decrease the inter-intra observer variability and save the human effort on labeling and classifying these images, a lot of research efforts have been devoted to the development of algorithms for biomedical images. Among such efforts, histology image classification is one of the most important areas due to its broad applications in pathological diagnosis such as cancer diagnosis. To improve classification accuracy, most of the previous work focuses on extracting more features and building algorithms for a specific task. This paper proposes a framework based on the novel and robust Collateral Representative Subspace Projection Modeling (C-RSPM) supervised classification model for general histology image classification. In the proposed framework, a cell image is first divided into 25 blocks to reduce the spatial complexity of computation, and one C-RSPM model is built on each block set which contains blocks in the same location from different images. For each testing image, our proposed framework first classifies each of its blocks using the C-RSPM classification model built for that block set, and then applies a multimodal late fusion algorithm with a weighted majority voting strategy to decide the final class label of the whole image. Experimenting using three-fold cross validation with three benchmark histology data sets shows that the proposed framework outperforms other well-known classifiers in the comparison and gives better results than the highest accuracy reported previously.",
keywords = "C-RSPM, Histology image classification, Multimodal fusion, Weighted majority voting algorithm",
author = "Tao Meng and Lin Lin and Mei-Ling Shyu and Chen, {Shu Ching}",
year = "2010",
month = "12",
day = "1",
doi = "10.1109/ISM.2010.29",
language = "English",
isbn = "9780769542171",
pages = "145--152",
booktitle = "Proceedings - 2010 IEEE International Symposium on Multimedia, ISM 2010",

}

TY - GEN

T1 - Histology image classification using supervised classification and multimodal fusion

AU - Meng, Tao

AU - Lin, Lin

AU - Shyu, Mei-Ling

AU - Chen, Shu Ching

PY - 2010/12/1

Y1 - 2010/12/1

N2 - The fast development of microscopy imaging techniques nowadays promotes the generation of a large amount of data. These data are very crucial not only for theoretical biomedical research but also for clinical usage. In order to decrease the inter-intra observer variability and save the human effort on labeling and classifying these images, a lot of research efforts have been devoted to the development of algorithms for biomedical images. Among such efforts, histology image classification is one of the most important areas due to its broad applications in pathological diagnosis such as cancer diagnosis. To improve classification accuracy, most of the previous work focuses on extracting more features and building algorithms for a specific task. This paper proposes a framework based on the novel and robust Collateral Representative Subspace Projection Modeling (C-RSPM) supervised classification model for general histology image classification. In the proposed framework, a cell image is first divided into 25 blocks to reduce the spatial complexity of computation, and one C-RSPM model is built on each block set which contains blocks in the same location from different images. For each testing image, our proposed framework first classifies each of its blocks using the C-RSPM classification model built for that block set, and then applies a multimodal late fusion algorithm with a weighted majority voting strategy to decide the final class label of the whole image. Experimenting using three-fold cross validation with three benchmark histology data sets shows that the proposed framework outperforms other well-known classifiers in the comparison and gives better results than the highest accuracy reported previously.

AB - The fast development of microscopy imaging techniques nowadays promotes the generation of a large amount of data. These data are very crucial not only for theoretical biomedical research but also for clinical usage. In order to decrease the inter-intra observer variability and save the human effort on labeling and classifying these images, a lot of research efforts have been devoted to the development of algorithms for biomedical images. Among such efforts, histology image classification is one of the most important areas due to its broad applications in pathological diagnosis such as cancer diagnosis. To improve classification accuracy, most of the previous work focuses on extracting more features and building algorithms for a specific task. This paper proposes a framework based on the novel and robust Collateral Representative Subspace Projection Modeling (C-RSPM) supervised classification model for general histology image classification. In the proposed framework, a cell image is first divided into 25 blocks to reduce the spatial complexity of computation, and one C-RSPM model is built on each block set which contains blocks in the same location from different images. For each testing image, our proposed framework first classifies each of its blocks using the C-RSPM classification model built for that block set, and then applies a multimodal late fusion algorithm with a weighted majority voting strategy to decide the final class label of the whole image. Experimenting using three-fold cross validation with three benchmark histology data sets shows that the proposed framework outperforms other well-known classifiers in the comparison and gives better results than the highest accuracy reported previously.

KW - C-RSPM

KW - Histology image classification

KW - Multimodal fusion

KW - Weighted majority voting algorithm

UR - http://www.scopus.com/inward/record.url?scp=79951733702&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=79951733702&partnerID=8YFLogxK

U2 - 10.1109/ISM.2010.29

DO - 10.1109/ISM.2010.29

M3 - Conference contribution

AN - SCOPUS:79951733702

SN - 9780769542171

SP - 145

EP - 152

BT - Proceedings - 2010 IEEE International Symposium on Multimedia, ISM 2010

ER -