Using spatial characteristics to aid automation of SOM segmentation of functional image data

Robert Grossman; Erzsébet Merényi; Patrick O’Driscoll

doi:10.1109/WSOM.2017.8020012

Using spatial characteristics to aid automation of SOM segmentation of functional image data

Robert Grossman, Erzsébet Merényi, Patrick O’Driscoll

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

2 Scopus citations

Abstract

We propose a new similarity measure, Combined Connectivity and Spatial Adjacency (CCSA), to be used in hierarchical agglomerative clustering (HAC) for automated segmentation of Self-Organizing Maps (SOMs, Kohonen [1]). The CCSA measure is specifically designed to assist segmentation of large, complex, functional image data by exploiting general spatial characteristics of such data. The proposed CCSA measure is constructed from two strong indicators of cluster structure: the degree of localization of data points in physical space and the degree of connectivity of SOM prototypes (as defined by Tasdemir and Merényi [2]). The new measure is expected to enhance cluster capture in large functional image data cubes such as hyperspectral imagery or fMRI brain images, where many relevant clusters exist with widely varying statistical properties and in complex relationships both in feature space and in physical (image) space. We demonstrate the effectiveness of our approach using the CCSA measure on progressively complex synthetic spatial data and on real fMRI brain data.

Original language	English (US)
Title of host publication	12th International Workshop on Self-Organizing Maps and Learning Vector Quantization, Clustering and Data Visualization, WSOM 2017 - Proceedings
Editors	Jean-Charles Lamirel, Madalina Olteanu, Marie Cottrell
Publisher	Institute of Electrical and Electronics Engineers Inc.
ISBN (Electronic)	9781509066384
DOIs	https://doi.org/10.1109/WSOM.2017.8020012
State	Published - Aug 29 2017
Event	12th International Workshop on Self-Organizing Maps and Learning Vector Quantization, Clustering and Data Visualization, WSOM 2017 - Nancy, France Duration: Jun 28 2017 → Jun 30 2017

Publication series

Name	12th International Workshop on Self-Organizing Maps and Learning Vector Quantization, Clustering and Data Visualization, WSOM 2017 - Proceedings

Other

Other	12th International Workshop on Self-Organizing Maps and Learning Vector Quantization, Clustering and Data Visualization, WSOM 2017
Country	France
City	Nancy
Period	6/28/17 → 6/30/17

ASJC Scopus subject areas

Artificial Intelligence
Computational Theory and Mathematics

Access to Document

10.1109/WSOM.2017.8020012

Cite this

Grossman, R., Merényi, E., & O’Driscoll, P. (2017). Using spatial characteristics to aid automation of SOM segmentation of functional image data. In J-C. Lamirel, M. Olteanu, & M. Cottrell (Eds.), 12th International Workshop on Self-Organizing Maps and Learning Vector Quantization, Clustering and Data Visualization, WSOM 2017 - Proceedings [8020012] (12th International Workshop on Self-Organizing Maps and Learning Vector Quantization, Clustering and Data Visualization, WSOM 2017 - Proceedings). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/WSOM.2017.8020012

Using spatial characteristics to aid automation of SOM segmentation of functional image data. / Grossman, Robert; Merényi, Erzsébet; O’Driscoll, Patrick.

12th International Workshop on Self-Organizing Maps and Learning Vector Quantization, Clustering and Data Visualization, WSOM 2017 - Proceedings. ed. / Jean-Charles Lamirel; Madalina Olteanu; Marie Cottrell. Institute of Electrical and Electronics Engineers Inc., 2017. 8020012 (12th International Workshop on Self-Organizing Maps and Learning Vector Quantization, Clustering and Data Visualization, WSOM 2017 - Proceedings).

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

Grossman, R, Merényi, E & O’Driscoll, P 2017, Using spatial characteristics to aid automation of SOM segmentation of functional image data. in J-C Lamirel, M Olteanu & M Cottrell (eds), 12th International Workshop on Self-Organizing Maps and Learning Vector Quantization, Clustering and Data Visualization, WSOM 2017 - Proceedings., 8020012, 12th International Workshop on Self-Organizing Maps and Learning Vector Quantization, Clustering and Data Visualization, WSOM 2017 - Proceedings, Institute of Electrical and Electronics Engineers Inc., 12th International Workshop on Self-Organizing Maps and Learning Vector Quantization, Clustering and Data Visualization, WSOM 2017, Nancy, France, 6/28/17. https://doi.org/10.1109/WSOM.2017.8020012

Grossman R, Merényi E, O’Driscoll P. Using spatial characteristics to aid automation of SOM segmentation of functional image data. In Lamirel J-C, Olteanu M, Cottrell M, editors, 12th International Workshop on Self-Organizing Maps and Learning Vector Quantization, Clustering and Data Visualization, WSOM 2017 - Proceedings. Institute of Electrical and Electronics Engineers Inc. 2017. 8020012. (12th International Workshop on Self-Organizing Maps and Learning Vector Quantization, Clustering and Data Visualization, WSOM 2017 - Proceedings). https://doi.org/10.1109/WSOM.2017.8020012

Grossman, Robert ; Merényi, Erzsébet ; O’Driscoll, Patrick. / Using spatial characteristics to aid automation of SOM segmentation of functional image data. 12th International Workshop on Self-Organizing Maps and Learning Vector Quantization, Clustering and Data Visualization, WSOM 2017 - Proceedings. editor / Jean-Charles Lamirel ; Madalina Olteanu ; Marie Cottrell. Institute of Electrical and Electronics Engineers Inc., 2017. (12th International Workshop on Self-Organizing Maps and Learning Vector Quantization, Clustering and Data Visualization, WSOM 2017 - Proceedings).

@inproceedings{bc51fcb71f6147c5975c89765227c9ed,

title = "Using spatial characteristics to aid automation of SOM segmentation of functional image data",

abstract = "We propose a new similarity measure, Combined Connectivity and Spatial Adjacency (CCSA), to be used in hierarchical agglomerative clustering (HAC) for automated segmentation of Self-Organizing Maps (SOMs, Kohonen [1]). The CCSA measure is specifically designed to assist segmentation of large, complex, functional image data by exploiting general spatial characteristics of such data. The proposed CCSA measure is constructed from two strong indicators of cluster structure: the degree of localization of data points in physical space and the degree of connectivity of SOM prototypes (as defined by Tasdemir and Mer{\'e}nyi [2]). The new measure is expected to enhance cluster capture in large functional image data cubes such as hyperspectral imagery or fMRI brain images, where many relevant clusters exist with widely varying statistical properties and in complex relationships both in feature space and in physical (image) space. We demonstrate the effectiveness of our approach using the CCSA measure on progressively complex synthetic spatial data and on real fMRI brain data.",

author = "Robert Grossman and Erzs{\'e}bet Mer{\'e}nyi and Patrick O{\textquoteright}Driscoll",

note = "Publisher Copyright: {\textcopyright} 2017 IEEE. Copyright: Copyright 2017 Elsevier B.V., All rights reserved.; 12th International Workshop on Self-Organizing Maps and Learning Vector Quantization, Clustering and Data Visualization, WSOM 2017 ; Conference date: 28-06-2017 Through 30-06-2017",

year = "2017",

month = aug,

day = "29",

doi = "10.1109/WSOM.2017.8020012",

language = "English (US)",

series = "12th International Workshop on Self-Organizing Maps and Learning Vector Quantization, Clustering and Data Visualization, WSOM 2017 - Proceedings",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

editor = "Jean-Charles Lamirel and Madalina Olteanu and Marie Cottrell",

booktitle = "12th International Workshop on Self-Organizing Maps and Learning Vector Quantization, Clustering and Data Visualization, WSOM 2017 - Proceedings",

address = "United States",

}

TY - GEN

T1 - Using spatial characteristics to aid automation of SOM segmentation of functional image data

AU - Grossman, Robert

AU - Merényi, Erzsébet

AU - O’Driscoll, Patrick

PY - 2017/8/29

Y1 - 2017/8/29

N2 - We propose a new similarity measure, Combined Connectivity and Spatial Adjacency (CCSA), to be used in hierarchical agglomerative clustering (HAC) for automated segmentation of Self-Organizing Maps (SOMs, Kohonen [1]). The CCSA measure is specifically designed to assist segmentation of large, complex, functional image data by exploiting general spatial characteristics of such data. The proposed CCSA measure is constructed from two strong indicators of cluster structure: the degree of localization of data points in physical space and the degree of connectivity of SOM prototypes (as defined by Tasdemir and Merényi [2]). The new measure is expected to enhance cluster capture in large functional image data cubes such as hyperspectral imagery or fMRI brain images, where many relevant clusters exist with widely varying statistical properties and in complex relationships both in feature space and in physical (image) space. We demonstrate the effectiveness of our approach using the CCSA measure on progressively complex synthetic spatial data and on real fMRI brain data.

AB - We propose a new similarity measure, Combined Connectivity and Spatial Adjacency (CCSA), to be used in hierarchical agglomerative clustering (HAC) for automated segmentation of Self-Organizing Maps (SOMs, Kohonen [1]). The CCSA measure is specifically designed to assist segmentation of large, complex, functional image data by exploiting general spatial characteristics of such data. The proposed CCSA measure is constructed from two strong indicators of cluster structure: the degree of localization of data points in physical space and the degree of connectivity of SOM prototypes (as defined by Tasdemir and Merényi [2]). The new measure is expected to enhance cluster capture in large functional image data cubes such as hyperspectral imagery or fMRI brain images, where many relevant clusters exist with widely varying statistical properties and in complex relationships both in feature space and in physical (image) space. We demonstrate the effectiveness of our approach using the CCSA measure on progressively complex synthetic spatial data and on real fMRI brain data.

UR - http://www.scopus.com/inward/record.url?scp=85031304219&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85031304219&partnerID=8YFLogxK

U2 - 10.1109/WSOM.2017.8020012

DO - 10.1109/WSOM.2017.8020012

M3 - Conference contribution

AN - SCOPUS:85031304219

T3 - 12th International Workshop on Self-Organizing Maps and Learning Vector Quantization, Clustering and Data Visualization, WSOM 2017 - Proceedings

BT - 12th International Workshop on Self-Organizing Maps and Learning Vector Quantization, Clustering and Data Visualization, WSOM 2017 - Proceedings

A2 - Lamirel, Jean-Charles

A2 - Olteanu, Madalina

A2 - Cottrell, Marie

PB - Institute of Electrical and Electronics Engineers Inc.

T2 - 12th International Workshop on Self-Organizing Maps and Learning Vector Quantization, Clustering and Data Visualization, WSOM 2017

Y2 - 28 June 2017 through 30 June 2017

ER -

Using spatial characteristics to aid automation of SOM segmentation of functional image data

Abstract

Publication series

Other

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this