Inter-cluster features for medical image classification

Siyamalan Manivannan; Ruixuan Wang; Emanuele Trucco

doi:10.1007/978-3-319-10443-0_44

Inter-cluster features for medical image classification

Siyamalan Manivannan, Ruixuan Wang, Emanuele Trucco

Computing

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

8 Citations (Scopus)

Abstract

Feature encoding plays an important role for medical image classification. Intra-cluster features such as bag of visual words have been widely used for feature encoding, which are based on the statistical information within each clusters of local features and therefore fail to capture the inter-cluster statistics, such as how the visual words co-occur in images. This paper proposes a new method to choose a subset of cluster pairs based on the idea of Latent Semantic Analysis (LSA) and proposes a new inter-cluster statistics which capture richer information than the traditional co-occurrence information. Since the cluster pairs are selected based on image patches rather than the whole images, the final representation also captures the local structures present in images. Experiments on medical datasets show that explicitly encoding inter-cluster statistics in addition to intra-cluster statistics significantly improves the classification performance, and adding the rich inter-cluster statistics performs better than the frequency based inter-cluster statistics.

Original language	English
Title of host publication	Medical Image Computing and Computer-Assisted Intervention – MICCAI 2014
Subtitle of host publication	17th International Conference, Boston, MA, USA, September 14-18, 2014, Proceedings, Part III
Editors	Polina Golland, Nobuhiko Hata, Christian Barillot, Joachim Hornegger, Robert Howe
Publisher	Springer International Publishing
Pages	345-352
Number of pages	8
ISBN (Electronic)	9783319104430
ISBN (Print)	9783319104423
DOIs	https://doi.org/10.1007/978-3-319-10443-0_44
Publication status	Published - 2014
Event	17th International Conference on Medical Image Computing and Computer-Assisted Intervention - Boston, MA., United States Duration: 14 Sept 2014 → 18 Sept 2014 http://miccai2014.org/index.html

Publication series

Name	Lecture notes in computer science
Publisher	Springer International Publishing
Volume	8675
ISSN (Print)	0302-9743

Conference

Conference	17th International Conference on Medical Image Computing and Computer-Assisted Intervention
Abbreviated title	MICCAI 2014
Country/Territory	United States
City	Boston, MA.
Period	14/09/14 → 18/09/14
Internet address	http://miccai2014.org/index.html

Access to Document

10.1007/978-3-319-10443-0_44

Cite this

Manivannan, S., Wang, R., & Trucco, E. (2014). Inter-cluster features for medical image classification. In P. Golland, N. Hata, C. Barillot, J. Hornegger, & R. Howe (Eds.), Medical Image Computing and Computer-Assisted Intervention – MICCAI 2014: 17th International Conference, Boston, MA, USA, September 14-18, 2014, Proceedings, Part III (pp. 345-352). (Lecture notes in computer science; Vol. 8675). Springer International Publishing. https://doi.org/10.1007/978-3-319-10443-0_44

Manivannan, Siyamalan ; Wang, Ruixuan ; Trucco, Emanuele. / Inter-cluster features for medical image classification. Medical Image Computing and Computer-Assisted Intervention – MICCAI 2014: 17th International Conference, Boston, MA, USA, September 14-18, 2014, Proceedings, Part III. editor / Polina Golland ; Nobuhiko Hata ; Christian Barillot ; Joachim Hornegger ; Robert Howe. Springer International Publishing, 2014. pp. 345-352 (Lecture notes in computer science).

@inproceedings{8d3f1ec645cf4c0890b9f99194f91ec6,

title = "Inter-cluster features for medical image classification",

abstract = "Feature encoding plays an important role for medical image classification. Intra-cluster features such as bag of visual words have been widely used for feature encoding, which are based on the statistical information within each clusters of local features and therefore fail to capture the inter-cluster statistics, such as how the visual words co-occur in images. This paper proposes a new method to choose a subset of cluster pairs based on the idea of Latent Semantic Analysis (LSA) and proposes a new inter-cluster statistics which capture richer information than the traditional co-occurrence information. Since the cluster pairs are selected based on image patches rather than the whole images, the final representation also captures the local structures present in images. Experiments on medical datasets show that explicitly encoding inter-cluster statistics in addition to intra-cluster statistics significantly improves the classification performance, and adding the rich inter-cluster statistics performs better than the frequency based inter-cluster statistics.",

author = "Siyamalan Manivannan and Ruixuan Wang and Emanuele Trucco",

year = "2014",

doi = "10.1007/978-3-319-10443-0_44",

language = "English",

isbn = "9783319104423",

series = "Lecture notes in computer science",

publisher = "Springer International Publishing",

pages = "345--352",

editor = "Polina Golland and Nobuhiko Hata and Barillot, {Christian } and Joachim Hornegger and Howe, {Robert }",

booktitle = "Medical Image Computing and Computer-Assisted Intervention – MICCAI 2014",

address = "Switzerland",

note = "17th International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2014 ; Conference date: 14-09-2014 Through 18-09-2014",

url = "http://miccai2014.org/index.html",

}

Manivannan, S, Wang, R & Trucco, E 2014, Inter-cluster features for medical image classification. in P Golland, N Hata, C Barillot, J Hornegger & R Howe (eds), Medical Image Computing and Computer-Assisted Intervention – MICCAI 2014: 17th International Conference, Boston, MA, USA, September 14-18, 2014, Proceedings, Part III. Lecture notes in computer science, vol. 8675, Springer International Publishing, pp. 345-352, 17th International Conference on Medical Image Computing and Computer-Assisted Intervention, Boston, MA., United States, 14/09/14. https://doi.org/10.1007/978-3-319-10443-0_44

Inter-cluster features for medical image classification. / Manivannan, Siyamalan; Wang, Ruixuan; Trucco, Emanuele.
Medical Image Computing and Computer-Assisted Intervention – MICCAI 2014: 17th International Conference, Boston, MA, USA, September 14-18, 2014, Proceedings, Part III. ed. / Polina Golland; Nobuhiko Hata; Christian Barillot; Joachim Hornegger; Robert Howe. Springer International Publishing, 2014. p. 345-352 (Lecture notes in computer science; Vol. 8675).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

TY - GEN

T1 - Inter-cluster features for medical image classification

AU - Manivannan, Siyamalan

AU - Wang, Ruixuan

AU - Trucco, Emanuele

PY - 2014

Y1 - 2014

N2 - Feature encoding plays an important role for medical image classification. Intra-cluster features such as bag of visual words have been widely used for feature encoding, which are based on the statistical information within each clusters of local features and therefore fail to capture the inter-cluster statistics, such as how the visual words co-occur in images. This paper proposes a new method to choose a subset of cluster pairs based on the idea of Latent Semantic Analysis (LSA) and proposes a new inter-cluster statistics which capture richer information than the traditional co-occurrence information. Since the cluster pairs are selected based on image patches rather than the whole images, the final representation also captures the local structures present in images. Experiments on medical datasets show that explicitly encoding inter-cluster statistics in addition to intra-cluster statistics significantly improves the classification performance, and adding the rich inter-cluster statistics performs better than the frequency based inter-cluster statistics.

AB - Feature encoding plays an important role for medical image classification. Intra-cluster features such as bag of visual words have been widely used for feature encoding, which are based on the statistical information within each clusters of local features and therefore fail to capture the inter-cluster statistics, such as how the visual words co-occur in images. This paper proposes a new method to choose a subset of cluster pairs based on the idea of Latent Semantic Analysis (LSA) and proposes a new inter-cluster statistics which capture richer information than the traditional co-occurrence information. Since the cluster pairs are selected based on image patches rather than the whole images, the final representation also captures the local structures present in images. Experiments on medical datasets show that explicitly encoding inter-cluster statistics in addition to intra-cluster statistics significantly improves the classification performance, and adding the rich inter-cluster statistics performs better than the frequency based inter-cluster statistics.

UR - http://www.scopus.com/inward/record.url?scp=84906974174&partnerID=8YFLogxK

U2 - 10.1007/978-3-319-10443-0_44

DO - 10.1007/978-3-319-10443-0_44

M3 - Conference contribution

AN - SCOPUS:84906974174

SN - 9783319104423

T3 - Lecture notes in computer science

SP - 345

EP - 352

BT - Medical Image Computing and Computer-Assisted Intervention – MICCAI 2014

A2 - Golland, Polina

A2 - Hata, Nobuhiko

A2 - Barillot, Christian

A2 - Hornegger, Joachim

A2 - Howe, Robert

PB - Springer International Publishing

T2 - 17th International Conference on Medical Image Computing and Computer-Assisted Intervention

Y2 - 14 September 2014 through 18 September 2014

ER -

Manivannan S, Wang R, Trucco E. Inter-cluster features for medical image classification. In Golland P, Hata N, Barillot C, Hornegger J, Howe R, editors, Medical Image Computing and Computer-Assisted Intervention – MICCAI 2014: 17th International Conference, Boston, MA, USA, September 14-18, 2014, Proceedings, Part III. Springer International Publishing. 2014. p. 345-352. (Lecture notes in computer science). doi: 10.1007/978-3-319-10443-0_44

Inter-cluster features for medical image classification

Abstract

Publication series

Conference

Access to Document

Other files and links

Fingerprint

Cite this