Inter-subject similarity guided brain network modeling for MCI diagnosis

Yu Zhang, Han Zhang, Xiaobo Chen, Mingxia Liu, Xiaofeng Zhu, Dinggang Shen

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

Abstract

Sparse representation-based brain network modeling, although popular, often results in relatively large inter-subject variability in network structures. This inevitably makes it difficult for inter-subject comparison, thus eventually deteriorating the generalization capability of personalized disease diagnosis. Accordingly, group sparse representation has been proposed to alleviate such limitation by jointly estimating connectivity weights for all subjects. However, the constructed brain networks based on this method often fail in providing satisfactory separability between the subjects from different groups (e.g., patients vs. normal controls), which will also affect the performance of computer-aided disease diagnosis. Based on the hypothesis that subjects from the same group should have larger similarity in their functional connectivity (FC) patterns than subjects from other groups, we propose an “inter-subject FC similarity-guided” group sparse network modeling method. In this method, we explicitly include the inter-subject FC similarity as a constraint to conduct group-wise FC network modeling, while retaining sufficient between-group differences in the resultant FC networks. This improves the separability of brain functional networks between different groups, thus facilitating better personalized brain disease diagnosis. Specifically, the inter-subject FC similarity is roughly estimated by comparing the Pearson’s correlation based FC patterns of each brain region to other regions for each pair of the subjects. Then, this is implemented as an additional weighting term to ensure the adequate inter-subject FC differences between the subjects from different groups. Of note, our method retains the group sparsity constraint to ensure the overall consistency of the resultant individual brain networks. Experimental results show that our method achieves a balanced trade-off by not only generating the individually consistent FC networks, but also effectively maintaining the necessary group difference, thereby significantly improving connectomics-based diagnosis for mild cognitive impairment (MCI).

Original languageEnglish
Title of host publicationMachine Learning in Medical Imaging - 8th International Workshop, MLMI 2017, Held in Conjunction with MICCAI 2017, Proceedings
PublisherSpringer Verlag
Pages168-175
Number of pages8
Volume10541 LNCS
ISBN (Print)9783319673882
DOIs
Publication statusPublished - 2017
Externally publishedYes
Event8th International Workshop on Machine Learning in Medical Imaging, MLMI 2017 held in conjunction with the 20th International Conference on Medical Image Computing and Computer Assisted Intervention, MICCAI 2017 - Quebec City, Canada
Duration: 2017 Sep 102017 Sep 10

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume10541 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Other

Other8th International Workshop on Machine Learning in Medical Imaging, MLMI 2017 held in conjunction with the 20th International Conference on Medical Image Computing and Computer Assisted Intervention, MICCAI 2017
CountryCanada
CityQuebec City
Period17/9/1017/9/10

Fingerprint

Network Modeling
Brain
Connectivity
Network Connectivity
Sparse Representation
Separability
Functional groups
Similarity
Pearson Correlation
Modeling Method
Sparsity
Network Structure
Weighting
Trade-offs
Sufficient

ASJC Scopus subject areas

  • Theoretical Computer Science
  • Computer Science(all)

Cite this

Zhang, Y., Zhang, H., Chen, X., Liu, M., Zhu, X., & Shen, D. (2017). Inter-subject similarity guided brain network modeling for MCI diagnosis. In Machine Learning in Medical Imaging - 8th International Workshop, MLMI 2017, Held in Conjunction with MICCAI 2017, Proceedings (Vol. 10541 LNCS, pp. 168-175). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 10541 LNCS). Springer Verlag. https://doi.org/10.1007/978-3-319-67389-9_20

Inter-subject similarity guided brain network modeling for MCI diagnosis. / Zhang, Yu; Zhang, Han; Chen, Xiaobo; Liu, Mingxia; Zhu, Xiaofeng; Shen, Dinggang.

Machine Learning in Medical Imaging - 8th International Workshop, MLMI 2017, Held in Conjunction with MICCAI 2017, Proceedings. Vol. 10541 LNCS Springer Verlag, 2017. p. 168-175 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 10541 LNCS).

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

Zhang, Y, Zhang, H, Chen, X, Liu, M, Zhu, X & Shen, D 2017, Inter-subject similarity guided brain network modeling for MCI diagnosis. in Machine Learning in Medical Imaging - 8th International Workshop, MLMI 2017, Held in Conjunction with MICCAI 2017, Proceedings. vol. 10541 LNCS, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 10541 LNCS, Springer Verlag, pp. 168-175, 8th International Workshop on Machine Learning in Medical Imaging, MLMI 2017 held in conjunction with the 20th International Conference on Medical Image Computing and Computer Assisted Intervention, MICCAI 2017, Quebec City, Canada, 17/9/10. https://doi.org/10.1007/978-3-319-67389-9_20
Zhang Y, Zhang H, Chen X, Liu M, Zhu X, Shen D. Inter-subject similarity guided brain network modeling for MCI diagnosis. In Machine Learning in Medical Imaging - 8th International Workshop, MLMI 2017, Held in Conjunction with MICCAI 2017, Proceedings. Vol. 10541 LNCS. Springer Verlag. 2017. p. 168-175. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). https://doi.org/10.1007/978-3-319-67389-9_20
Zhang, Yu ; Zhang, Han ; Chen, Xiaobo ; Liu, Mingxia ; Zhu, Xiaofeng ; Shen, Dinggang. / Inter-subject similarity guided brain network modeling for MCI diagnosis. Machine Learning in Medical Imaging - 8th International Workshop, MLMI 2017, Held in Conjunction with MICCAI 2017, Proceedings. Vol. 10541 LNCS Springer Verlag, 2017. pp. 168-175 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
@inproceedings{a645835a721f423f863676fb97d138a5,
title = "Inter-subject similarity guided brain network modeling for MCI diagnosis",
abstract = "Sparse representation-based brain network modeling, although popular, often results in relatively large inter-subject variability in network structures. This inevitably makes it difficult for inter-subject comparison, thus eventually deteriorating the generalization capability of personalized disease diagnosis. Accordingly, group sparse representation has been proposed to alleviate such limitation by jointly estimating connectivity weights for all subjects. However, the constructed brain networks based on this method often fail in providing satisfactory separability between the subjects from different groups (e.g., patients vs. normal controls), which will also affect the performance of computer-aided disease diagnosis. Based on the hypothesis that subjects from the same group should have larger similarity in their functional connectivity (FC) patterns than subjects from other groups, we propose an “inter-subject FC similarity-guided” group sparse network modeling method. In this method, we explicitly include the inter-subject FC similarity as a constraint to conduct group-wise FC network modeling, while retaining sufficient between-group differences in the resultant FC networks. This improves the separability of brain functional networks between different groups, thus facilitating better personalized brain disease diagnosis. Specifically, the inter-subject FC similarity is roughly estimated by comparing the Pearson’s correlation based FC patterns of each brain region to other regions for each pair of the subjects. Then, this is implemented as an additional weighting term to ensure the adequate inter-subject FC differences between the subjects from different groups. Of note, our method retains the group sparsity constraint to ensure the overall consistency of the resultant individual brain networks. Experimental results show that our method achieves a balanced trade-off by not only generating the individually consistent FC networks, but also effectively maintaining the necessary group difference, thereby significantly improving connectomics-based diagnosis for mild cognitive impairment (MCI).",
author = "Yu Zhang and Han Zhang and Xiaobo Chen and Mingxia Liu and Xiaofeng Zhu and Dinggang Shen",
year = "2017",
doi = "10.1007/978-3-319-67389-9_20",
language = "English",
isbn = "9783319673882",
volume = "10541 LNCS",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
publisher = "Springer Verlag",
pages = "168--175",
booktitle = "Machine Learning in Medical Imaging - 8th International Workshop, MLMI 2017, Held in Conjunction with MICCAI 2017, Proceedings",

}

TY - GEN

T1 - Inter-subject similarity guided brain network modeling for MCI diagnosis

AU - Zhang, Yu

AU - Zhang, Han

AU - Chen, Xiaobo

AU - Liu, Mingxia

AU - Zhu, Xiaofeng

AU - Shen, Dinggang

PY - 2017

Y1 - 2017

N2 - Sparse representation-based brain network modeling, although popular, often results in relatively large inter-subject variability in network structures. This inevitably makes it difficult for inter-subject comparison, thus eventually deteriorating the generalization capability of personalized disease diagnosis. Accordingly, group sparse representation has been proposed to alleviate such limitation by jointly estimating connectivity weights for all subjects. However, the constructed brain networks based on this method often fail in providing satisfactory separability between the subjects from different groups (e.g., patients vs. normal controls), which will also affect the performance of computer-aided disease diagnosis. Based on the hypothesis that subjects from the same group should have larger similarity in their functional connectivity (FC) patterns than subjects from other groups, we propose an “inter-subject FC similarity-guided” group sparse network modeling method. In this method, we explicitly include the inter-subject FC similarity as a constraint to conduct group-wise FC network modeling, while retaining sufficient between-group differences in the resultant FC networks. This improves the separability of brain functional networks between different groups, thus facilitating better personalized brain disease diagnosis. Specifically, the inter-subject FC similarity is roughly estimated by comparing the Pearson’s correlation based FC patterns of each brain region to other regions for each pair of the subjects. Then, this is implemented as an additional weighting term to ensure the adequate inter-subject FC differences between the subjects from different groups. Of note, our method retains the group sparsity constraint to ensure the overall consistency of the resultant individual brain networks. Experimental results show that our method achieves a balanced trade-off by not only generating the individually consistent FC networks, but also effectively maintaining the necessary group difference, thereby significantly improving connectomics-based diagnosis for mild cognitive impairment (MCI).

AB - Sparse representation-based brain network modeling, although popular, often results in relatively large inter-subject variability in network structures. This inevitably makes it difficult for inter-subject comparison, thus eventually deteriorating the generalization capability of personalized disease diagnosis. Accordingly, group sparse representation has been proposed to alleviate such limitation by jointly estimating connectivity weights for all subjects. However, the constructed brain networks based on this method often fail in providing satisfactory separability between the subjects from different groups (e.g., patients vs. normal controls), which will also affect the performance of computer-aided disease diagnosis. Based on the hypothesis that subjects from the same group should have larger similarity in their functional connectivity (FC) patterns than subjects from other groups, we propose an “inter-subject FC similarity-guided” group sparse network modeling method. In this method, we explicitly include the inter-subject FC similarity as a constraint to conduct group-wise FC network modeling, while retaining sufficient between-group differences in the resultant FC networks. This improves the separability of brain functional networks between different groups, thus facilitating better personalized brain disease diagnosis. Specifically, the inter-subject FC similarity is roughly estimated by comparing the Pearson’s correlation based FC patterns of each brain region to other regions for each pair of the subjects. Then, this is implemented as an additional weighting term to ensure the adequate inter-subject FC differences between the subjects from different groups. Of note, our method retains the group sparsity constraint to ensure the overall consistency of the resultant individual brain networks. Experimental results show that our method achieves a balanced trade-off by not only generating the individually consistent FC networks, but also effectively maintaining the necessary group difference, thereby significantly improving connectomics-based diagnosis for mild cognitive impairment (MCI).

UR - http://www.scopus.com/inward/record.url?scp=85029710432&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85029710432&partnerID=8YFLogxK

U2 - 10.1007/978-3-319-67389-9_20

DO - 10.1007/978-3-319-67389-9_20

M3 - Conference contribution

AN - SCOPUS:85029710432

SN - 9783319673882

VL - 10541 LNCS

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

SP - 168

EP - 175

BT - Machine Learning in Medical Imaging - 8th International Workshop, MLMI 2017, Held in Conjunction with MICCAI 2017, Proceedings

PB - Springer Verlag

ER -