Fiber modeling and clustering based on neuroanatomical features

Qian Wang, Pew Thian Yap, Guorong Wu, Dinggang Shen

Research output: Contribution to journalArticle

1 Citation (Scopus)

Abstract

DTI tractography allows unprecedented understanding of brain neural connectivity in-vivo by capturing water diffusion patterns in brain white-matter microstructures. However, tractography algorithms often output hundreds of thousands of fibers, rendering the computation needed for subsequent data analysis intractable. A remedy is to group the fibers into bundles using fiber clustering techniques. Most existing fiber clustering methods, however, rely on fiber geometrical information only by viewing fibers as curves in the 3D Euclidean space. The important neuroanatomical aspect of the fibers is mostly ignored. In this paper, neuroanatomical information is encapsulated in a feature vector called the associativity vector, which functions as the "fingerprint" for each fiber and depicts the connectivity of the fiber with respect to individual anatomies. Using the associativity vectors of fibers, we model the fibers as observations sampled from multivariate Gaussian mixtures in the feature space. An expectation-maximization clustering approach is then employed to group the fibers into 16 major bundles. Experimental results indicate that the proposed method groups the fibers into anatomically meaningful bundles, which are highly consistent across subjects.

Original languageEnglish
Pages (from-to)17-24
Number of pages8
JournalLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume6892 LNCS
Issue numberPART 2
DOIs
Publication statusPublished - 2011 Oct 11
Externally publishedYes

Fingerprint

Clustering
Fiber
Fibers
Modeling
Associativity
Bundle
Connectivity
Brain
Gaussian Mixture
Expectation Maximization
Fiber Bundle
Anatomy
Fingerprint
Feature Space
Feature Vector
Clustering Methods
Rendering
Euclidean space
Microstructure
Data analysis

ASJC Scopus subject areas

  • Computer Science(all)
  • Theoretical Computer Science

Cite this

Fiber modeling and clustering based on neuroanatomical features. / Wang, Qian; Yap, Pew Thian; Wu, Guorong; Shen, Dinggang.

In: Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), Vol. 6892 LNCS, No. PART 2, 11.10.2011, p. 17-24.

Research output: Contribution to journalArticle

@article{ad0f95b8dec147deb96b284dcddf5a26,
title = "Fiber modeling and clustering based on neuroanatomical features",
abstract = "DTI tractography allows unprecedented understanding of brain neural connectivity in-vivo by capturing water diffusion patterns in brain white-matter microstructures. However, tractography algorithms often output hundreds of thousands of fibers, rendering the computation needed for subsequent data analysis intractable. A remedy is to group the fibers into bundles using fiber clustering techniques. Most existing fiber clustering methods, however, rely on fiber geometrical information only by viewing fibers as curves in the 3D Euclidean space. The important neuroanatomical aspect of the fibers is mostly ignored. In this paper, neuroanatomical information is encapsulated in a feature vector called the associativity vector, which functions as the {"}fingerprint{"} for each fiber and depicts the connectivity of the fiber with respect to individual anatomies. Using the associativity vectors of fibers, we model the fibers as observations sampled from multivariate Gaussian mixtures in the feature space. An expectation-maximization clustering approach is then employed to group the fibers into 16 major bundles. Experimental results indicate that the proposed method groups the fibers into anatomically meaningful bundles, which are highly consistent across subjects.",
author = "Qian Wang and Yap, {Pew Thian} and Guorong Wu and Dinggang Shen",
year = "2011",
month = "10",
day = "11",
doi = "10.1007/978-3-642-23629-7_3",
language = "English",
volume = "6892 LNCS",
pages = "17--24",
journal = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
issn = "0302-9743",
publisher = "Springer Verlag",
number = "PART 2",

}

TY - JOUR

T1 - Fiber modeling and clustering based on neuroanatomical features

AU - Wang, Qian

AU - Yap, Pew Thian

AU - Wu, Guorong

AU - Shen, Dinggang

PY - 2011/10/11

Y1 - 2011/10/11

N2 - DTI tractography allows unprecedented understanding of brain neural connectivity in-vivo by capturing water diffusion patterns in brain white-matter microstructures. However, tractography algorithms often output hundreds of thousands of fibers, rendering the computation needed for subsequent data analysis intractable. A remedy is to group the fibers into bundles using fiber clustering techniques. Most existing fiber clustering methods, however, rely on fiber geometrical information only by viewing fibers as curves in the 3D Euclidean space. The important neuroanatomical aspect of the fibers is mostly ignored. In this paper, neuroanatomical information is encapsulated in a feature vector called the associativity vector, which functions as the "fingerprint" for each fiber and depicts the connectivity of the fiber with respect to individual anatomies. Using the associativity vectors of fibers, we model the fibers as observations sampled from multivariate Gaussian mixtures in the feature space. An expectation-maximization clustering approach is then employed to group the fibers into 16 major bundles. Experimental results indicate that the proposed method groups the fibers into anatomically meaningful bundles, which are highly consistent across subjects.

AB - DTI tractography allows unprecedented understanding of brain neural connectivity in-vivo by capturing water diffusion patterns in brain white-matter microstructures. However, tractography algorithms often output hundreds of thousands of fibers, rendering the computation needed for subsequent data analysis intractable. A remedy is to group the fibers into bundles using fiber clustering techniques. Most existing fiber clustering methods, however, rely on fiber geometrical information only by viewing fibers as curves in the 3D Euclidean space. The important neuroanatomical aspect of the fibers is mostly ignored. In this paper, neuroanatomical information is encapsulated in a feature vector called the associativity vector, which functions as the "fingerprint" for each fiber and depicts the connectivity of the fiber with respect to individual anatomies. Using the associativity vectors of fibers, we model the fibers as observations sampled from multivariate Gaussian mixtures in the feature space. An expectation-maximization clustering approach is then employed to group the fibers into 16 major bundles. Experimental results indicate that the proposed method groups the fibers into anatomically meaningful bundles, which are highly consistent across subjects.

UR - http://www.scopus.com/inward/record.url?scp=80053555186&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=80053555186&partnerID=8YFLogxK

U2 - 10.1007/978-3-642-23629-7_3

DO - 10.1007/978-3-642-23629-7_3

M3 - Article

AN - SCOPUS:80053555186

VL - 6892 LNCS

SP - 17

EP - 24

JO - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

JF - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SN - 0302-9743

IS - PART 2

ER -