Estimation of brain network atlases using diffusive-shrinking graphs: Application to developing brains

Islem Rekik, Gang Li, Weili Lin, Dinggang Shen

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

5 Citations (Scopus)

Abstract

Many methods have been developed to spatially normalize a population of brain images for estimating a mean image as a populationaverage atlas. However, methods for deriving a network atlas from a set of brain networks sitting on a complex manifold are still absent. Learning how to average brain networks across subjects constitutes a key step in creating a reliable mean representation of a population of brain networks, which can be used to spot abnormal deviations from the healthy network atlas. In this work, we propose a novel network atlas estimation framework, which guarantees that the produced network atlas is clean (for tuning down noisy measurements) and well-centered (for being optimally close to all subjects and representing the individual traits of each subject in the population). Specifically, for a population of brain networks, we first build a tensor, where each of its frontal-views (i.e., frontal matrices) represents a connectivity network matrix of a single subject in the population. Then, we use tensor robust principal component analysis for jointly denoising all subjects’ networks through cleaving a sparse noisy network population tensor from a clean low-rank network tensor. Second, we build a graph where each node represents a frontal-view of the unfolded clean tensor (network), to leverage the local manifold structure of these networks when fusing them. Specifically, we progressively shrink the graph of networks towards the centered mean network atlas through non-linear diffusion along the local neighbors of each of its nodes. Our evaluation on the developing functional and morphological brain networks at 1, 3, 6, 9 and 12 months of age has showed a better centeredness of our network atlases, in comparison with the baseline network fusion method. Further cleaning of the population of networks produces even more centered atlases, especially for the noisy functional connectivity networks.

Original languageEnglish
Title of host publicationInformation Processing in Medical Imaging - 25th International Conference, IPMI 2017, Proceedings
PublisherSpringer Verlag
Pages385-397
Number of pages13
Volume10265 LNCS
ISBN (Print)9783319590493
DOIs
Publication statusPublished - 2017
Externally publishedYes
Event25th International Conference on Information Processing in Medical Imaging, IPMI 2017 - Boone, United States
Duration: 2017 Jun 252017 Jun 30

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume10265 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Other

Other25th International Conference on Information Processing in Medical Imaging, IPMI 2017
CountryUnited States
CityBoone
Period17/6/2517/6/30

Fingerprint

Atlas
Shrinking
Brain
Tensors
Graph in graph theory
Tensor
Principal component analysis
Cleaning
Fusion reactions
Tuning
Network Connectivity
Normalize
Nonlinear Diffusion
Complex Manifolds

ASJC Scopus subject areas

  • Theoretical Computer Science
  • Computer Science(all)

Cite this

Rekik, I., Li, G., Lin, W., & Shen, D. (2017). Estimation of brain network atlases using diffusive-shrinking graphs: Application to developing brains. In Information Processing in Medical Imaging - 25th International Conference, IPMI 2017, Proceedings (Vol. 10265 LNCS, pp. 385-397). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 10265 LNCS). Springer Verlag. https://doi.org/10.1007/978-3-319-59050-9_31

Estimation of brain network atlases using diffusive-shrinking graphs : Application to developing brains. / Rekik, Islem; Li, Gang; Lin, Weili; Shen, Dinggang.

Information Processing in Medical Imaging - 25th International Conference, IPMI 2017, Proceedings. Vol. 10265 LNCS Springer Verlag, 2017. p. 385-397 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 10265 LNCS).

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

Rekik, I, Li, G, Lin, W & Shen, D 2017, Estimation of brain network atlases using diffusive-shrinking graphs: Application to developing brains. in Information Processing in Medical Imaging - 25th International Conference, IPMI 2017, Proceedings. vol. 10265 LNCS, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 10265 LNCS, Springer Verlag, pp. 385-397, 25th International Conference on Information Processing in Medical Imaging, IPMI 2017, Boone, United States, 17/6/25. https://doi.org/10.1007/978-3-319-59050-9_31
Rekik I, Li G, Lin W, Shen D. Estimation of brain network atlases using diffusive-shrinking graphs: Application to developing brains. In Information Processing in Medical Imaging - 25th International Conference, IPMI 2017, Proceedings. Vol. 10265 LNCS. Springer Verlag. 2017. p. 385-397. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). https://doi.org/10.1007/978-3-319-59050-9_31
Rekik, Islem ; Li, Gang ; Lin, Weili ; Shen, Dinggang. / Estimation of brain network atlases using diffusive-shrinking graphs : Application to developing brains. Information Processing in Medical Imaging - 25th International Conference, IPMI 2017, Proceedings. Vol. 10265 LNCS Springer Verlag, 2017. pp. 385-397 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
@inproceedings{af84ea5d8f934e69b5a3a086a1222def,
title = "Estimation of brain network atlases using diffusive-shrinking graphs: Application to developing brains",
abstract = "Many methods have been developed to spatially normalize a population of brain images for estimating a mean image as a populationaverage atlas. However, methods for deriving a network atlas from a set of brain networks sitting on a complex manifold are still absent. Learning how to average brain networks across subjects constitutes a key step in creating a reliable mean representation of a population of brain networks, which can be used to spot abnormal deviations from the healthy network atlas. In this work, we propose a novel network atlas estimation framework, which guarantees that the produced network atlas is clean (for tuning down noisy measurements) and well-centered (for being optimally close to all subjects and representing the individual traits of each subject in the population). Specifically, for a population of brain networks, we first build a tensor, where each of its frontal-views (i.e., frontal matrices) represents a connectivity network matrix of a single subject in the population. Then, we use tensor robust principal component analysis for jointly denoising all subjects’ networks through cleaving a sparse noisy network population tensor from a clean low-rank network tensor. Second, we build a graph where each node represents a frontal-view of the unfolded clean tensor (network), to leverage the local manifold structure of these networks when fusing them. Specifically, we progressively shrink the graph of networks towards the centered mean network atlas through non-linear diffusion along the local neighbors of each of its nodes. Our evaluation on the developing functional and morphological brain networks at 1, 3, 6, 9 and 12 months of age has showed a better centeredness of our network atlases, in comparison with the baseline network fusion method. Further cleaning of the population of networks produces even more centered atlases, especially for the noisy functional connectivity networks.",
author = "Islem Rekik and Gang Li and Weili Lin and Dinggang Shen",
year = "2017",
doi = "10.1007/978-3-319-59050-9_31",
language = "English",
isbn = "9783319590493",
volume = "10265 LNCS",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
publisher = "Springer Verlag",
pages = "385--397",
booktitle = "Information Processing in Medical Imaging - 25th International Conference, IPMI 2017, Proceedings",

}

TY - GEN

T1 - Estimation of brain network atlases using diffusive-shrinking graphs

T2 - Application to developing brains

AU - Rekik, Islem

AU - Li, Gang

AU - Lin, Weili

AU - Shen, Dinggang

PY - 2017

Y1 - 2017

N2 - Many methods have been developed to spatially normalize a population of brain images for estimating a mean image as a populationaverage atlas. However, methods for deriving a network atlas from a set of brain networks sitting on a complex manifold are still absent. Learning how to average brain networks across subjects constitutes a key step in creating a reliable mean representation of a population of brain networks, which can be used to spot abnormal deviations from the healthy network atlas. In this work, we propose a novel network atlas estimation framework, which guarantees that the produced network atlas is clean (for tuning down noisy measurements) and well-centered (for being optimally close to all subjects and representing the individual traits of each subject in the population). Specifically, for a population of brain networks, we first build a tensor, where each of its frontal-views (i.e., frontal matrices) represents a connectivity network matrix of a single subject in the population. Then, we use tensor robust principal component analysis for jointly denoising all subjects’ networks through cleaving a sparse noisy network population tensor from a clean low-rank network tensor. Second, we build a graph where each node represents a frontal-view of the unfolded clean tensor (network), to leverage the local manifold structure of these networks when fusing them. Specifically, we progressively shrink the graph of networks towards the centered mean network atlas through non-linear diffusion along the local neighbors of each of its nodes. Our evaluation on the developing functional and morphological brain networks at 1, 3, 6, 9 and 12 months of age has showed a better centeredness of our network atlases, in comparison with the baseline network fusion method. Further cleaning of the population of networks produces even more centered atlases, especially for the noisy functional connectivity networks.

AB - Many methods have been developed to spatially normalize a population of brain images for estimating a mean image as a populationaverage atlas. However, methods for deriving a network atlas from a set of brain networks sitting on a complex manifold are still absent. Learning how to average brain networks across subjects constitutes a key step in creating a reliable mean representation of a population of brain networks, which can be used to spot abnormal deviations from the healthy network atlas. In this work, we propose a novel network atlas estimation framework, which guarantees that the produced network atlas is clean (for tuning down noisy measurements) and well-centered (for being optimally close to all subjects and representing the individual traits of each subject in the population). Specifically, for a population of brain networks, we first build a tensor, where each of its frontal-views (i.e., frontal matrices) represents a connectivity network matrix of a single subject in the population. Then, we use tensor robust principal component analysis for jointly denoising all subjects’ networks through cleaving a sparse noisy network population tensor from a clean low-rank network tensor. Second, we build a graph where each node represents a frontal-view of the unfolded clean tensor (network), to leverage the local manifold structure of these networks when fusing them. Specifically, we progressively shrink the graph of networks towards the centered mean network atlas through non-linear diffusion along the local neighbors of each of its nodes. Our evaluation on the developing functional and morphological brain networks at 1, 3, 6, 9 and 12 months of age has showed a better centeredness of our network atlases, in comparison with the baseline network fusion method. Further cleaning of the population of networks produces even more centered atlases, especially for the noisy functional connectivity networks.

UR - http://www.scopus.com/inward/record.url?scp=85020500105&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85020500105&partnerID=8YFLogxK

U2 - 10.1007/978-3-319-59050-9_31

DO - 10.1007/978-3-319-59050-9_31

M3 - Conference contribution

AN - SCOPUS:85020500105

SN - 9783319590493

VL - 10265 LNCS

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 385

EP - 397

BT - Information Processing in Medical Imaging - 25th International Conference, IPMI 2017, Proceedings

PB - Springer Verlag

ER -