Deep Modeling of Growth Trajectories for Longitudinal Prediction of Missing Infant Cortical Surfaces

Peirong Liu; Zhengwang Wu; Gang Li; Pew Thian Yap; Dinggang Shen

doi:10.1007/978-3-030-20351-1_21

Deep Modeling of Growth Trajectories for Longitudinal Prediction of Missing Infant Cortical Surfaces

Peirong Liu, Zhengwang Wu, Gang Li, Pew Thian Yap, Dinggang Shen

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

5 Citations (Scopus)

Abstract

Charting cortical growth trajectories is of paramount importance for understanding brain development. However, such analysis necessitates the collection of longitudinal data, which can be challenging due to subject dropouts and failed scans. In this paper, we will introduce a method for longitudinal prediction of cortical surfaces using a spatial graph convolutional neural network (GCNN), which extends conventional CNNs from Euclidean to curved manifolds. The proposed method is designed to model the cortical growth trajectories and jointly predict inner and outer cortical surfaces at multiple time points. Adopting a binary flag in loss calculation to deal with missing data, we fully utilize all available cortical surfaces for training our deep learning model, without requiring a complete collection of longitudinal data. Predicting the surfaces directly allows cortical attributes such as cortical thickness, curvature, and convexity to be computed for subsequent analysis. We will demonstrate with experimental results that our method is capable of capturing the nonlinearity of spatiotemporal cortical growth patterns and can predict cortical surfaces with improved accuracy.

Original language	English
Title of host publication	Information Processing in Medical Imaging - 26th International Conference, IPMI 2019, Proceedings
Editors	Albert C.S. Chung, Siqi Bao, James C. Gee, Paul A. Yushkevich
Publisher	Springer Verlag
Pages	277-288
Number of pages	12
ISBN (Print)	9783030203504
DOIs	https://doi.org/10.1007/978-3-030-20351-1_21
Publication status	Published - 2019
Event	26th International Conference on Information Processing in Medical Imaging, IPMI 2019 - Hong Kong, China Duration: 2019 Jun 2 → 2019 Jun 7

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	11492 LNCS
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	26th International Conference on Information Processing in Medical Imaging, IPMI 2019
Country/Territory	China
City	Hong Kong
Period	19/6/2 → 19/6/7

Bibliographical note

Publisher Copyright:
© 2019, Springer Nature Switzerland AG.

Keywords

Graph Convolutional Neural Networks
Infant cortical surfaces
Longitudinal prediction
Missing data
Shape Analysis

ASJC Scopus subject areas

Theoretical Computer Science
General Computer Science

Access to Document

10.1007/978-3-030-20351-1_21

Cite this

Liu, P., Wu, Z., Li, G., Yap, P. T., & Shen, D. (2019). Deep Modeling of Growth Trajectories for Longitudinal Prediction of Missing Infant Cortical Surfaces. In A. C. S. Chung, S. Bao, J. C. Gee, & P. A. Yushkevich (Eds.), Information Processing in Medical Imaging - 26th International Conference, IPMI 2019, Proceedings (pp. 277-288). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 11492 LNCS). Springer Verlag. https://doi.org/10.1007/978-3-030-20351-1_21

Deep Modeling of Growth Trajectories for Longitudinal Prediction of Missing Infant Cortical Surfaces. / Liu, Peirong; Wu, Zhengwang; Li, Gang et al.
Information Processing in Medical Imaging - 26th International Conference, IPMI 2019, Proceedings. ed. / Albert C.S. Chung; Siqi Bao; James C. Gee; Paul A. Yushkevich. Springer Verlag, 2019. p. 277-288 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 11492 LNCS).

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

Liu, P, Wu, Z, Li, G, Yap, PT & Shen, D 2019, Deep Modeling of Growth Trajectories for Longitudinal Prediction of Missing Infant Cortical Surfaces. in ACS Chung, S Bao, JC Gee & PA Yushkevich (eds), Information Processing in Medical Imaging - 26th International Conference, IPMI 2019, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 11492 LNCS, Springer Verlag, pp. 277-288, 26th International Conference on Information Processing in Medical Imaging, IPMI 2019, Hong Kong, China, 19/6/2. https://doi.org/10.1007/978-3-030-20351-1_21

Liu P, Wu Z, Li G, Yap PT, Shen D. Deep Modeling of Growth Trajectories for Longitudinal Prediction of Missing Infant Cortical Surfaces. In Chung ACS, Bao S, Gee JC, Yushkevich PA, editors, Information Processing in Medical Imaging - 26th International Conference, IPMI 2019, Proceedings. Springer Verlag. 2019. p. 277-288. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-030-20351-1_21

Liu, Peirong ; Wu, Zhengwang ; Li, Gang et al. / Deep Modeling of Growth Trajectories for Longitudinal Prediction of Missing Infant Cortical Surfaces. Information Processing in Medical Imaging - 26th International Conference, IPMI 2019, Proceedings. editor / Albert C.S. Chung ; Siqi Bao ; James C. Gee ; Paul A. Yushkevich. Springer Verlag, 2019. pp. 277-288 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{45242b0f3972416e9c2bddcfe257e849,

title = "Deep Modeling of Growth Trajectories for Longitudinal Prediction of Missing Infant Cortical Surfaces",

abstract = "Charting cortical growth trajectories is of paramount importance for understanding brain development. However, such analysis necessitates the collection of longitudinal data, which can be challenging due to subject dropouts and failed scans. In this paper, we will introduce a method for longitudinal prediction of cortical surfaces using a spatial graph convolutional neural network (GCNN), which extends conventional CNNs from Euclidean to curved manifolds. The proposed method is designed to model the cortical growth trajectories and jointly predict inner and outer cortical surfaces at multiple time points. Adopting a binary flag in loss calculation to deal with missing data, we fully utilize all available cortical surfaces for training our deep learning model, without requiring a complete collection of longitudinal data. Predicting the surfaces directly allows cortical attributes such as cortical thickness, curvature, and convexity to be computed for subsequent analysis. We will demonstrate with experimental results that our method is capable of capturing the nonlinearity of spatiotemporal cortical growth patterns and can predict cortical surfaces with improved accuracy.",

keywords = "Graph Convolutional Neural Networks, Infant cortical surfaces, Longitudinal prediction, Missing data, Shape Analysis",

author = "Peirong Liu and Zhengwang Wu and Gang Li and Yap, {Pew Thian} and Dinggang Shen",

note = "Publisher Copyright: {\textcopyright} 2019, Springer Nature Switzerland AG.; 26th International Conference on Information Processing in Medical Imaging, IPMI 2019 ; Conference date: 02-06-2019 Through 07-06-2019",

year = "2019",

doi = "10.1007/978-3-030-20351-1_21",

language = "English",

isbn = "9783030203504",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

publisher = "Springer Verlag",

pages = "277--288",

editor = "Chung, {Albert C.S.} and Siqi Bao and Gee, {James C.} and Yushkevich, {Paul A.}",

booktitle = "Information Processing in Medical Imaging - 26th International Conference, IPMI 2019, Proceedings",

}

TY - GEN

T1 - Deep Modeling of Growth Trajectories for Longitudinal Prediction of Missing Infant Cortical Surfaces

AU - Liu, Peirong

AU - Wu, Zhengwang

AU - Li, Gang

AU - Yap, Pew Thian

AU - Shen, Dinggang

PY - 2019

Y1 - 2019

N2 - Charting cortical growth trajectories is of paramount importance for understanding brain development. However, such analysis necessitates the collection of longitudinal data, which can be challenging due to subject dropouts and failed scans. In this paper, we will introduce a method for longitudinal prediction of cortical surfaces using a spatial graph convolutional neural network (GCNN), which extends conventional CNNs from Euclidean to curved manifolds. The proposed method is designed to model the cortical growth trajectories and jointly predict inner and outer cortical surfaces at multiple time points. Adopting a binary flag in loss calculation to deal with missing data, we fully utilize all available cortical surfaces for training our deep learning model, without requiring a complete collection of longitudinal data. Predicting the surfaces directly allows cortical attributes such as cortical thickness, curvature, and convexity to be computed for subsequent analysis. We will demonstrate with experimental results that our method is capable of capturing the nonlinearity of spatiotemporal cortical growth patterns and can predict cortical surfaces with improved accuracy.

AB - Charting cortical growth trajectories is of paramount importance for understanding brain development. However, such analysis necessitates the collection of longitudinal data, which can be challenging due to subject dropouts and failed scans. In this paper, we will introduce a method for longitudinal prediction of cortical surfaces using a spatial graph convolutional neural network (GCNN), which extends conventional CNNs from Euclidean to curved manifolds. The proposed method is designed to model the cortical growth trajectories and jointly predict inner and outer cortical surfaces at multiple time points. Adopting a binary flag in loss calculation to deal with missing data, we fully utilize all available cortical surfaces for training our deep learning model, without requiring a complete collection of longitudinal data. Predicting the surfaces directly allows cortical attributes such as cortical thickness, curvature, and convexity to be computed for subsequent analysis. We will demonstrate with experimental results that our method is capable of capturing the nonlinearity of spatiotemporal cortical growth patterns and can predict cortical surfaces with improved accuracy.

KW - Graph Convolutional Neural Networks

KW - Infant cortical surfaces

KW - Longitudinal prediction

KW - Missing data

KW - Shape Analysis

UR - http://www.scopus.com/inward/record.url?scp=85066148992&partnerID=8YFLogxK

U2 - 10.1007/978-3-030-20351-1_21

DO - 10.1007/978-3-030-20351-1_21

M3 - Conference contribution

AN - SCOPUS:85066148992

SN - 9783030203504

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

SP - 277

EP - 288

BT - Information Processing in Medical Imaging - 26th International Conference, IPMI 2019, Proceedings

A2 - Chung, Albert C.S.

A2 - Bao, Siqi

A2 - Gee, James C.

A2 - Yushkevich, Paul A.

PB - Springer Verlag

T2 - 26th International Conference on Information Processing in Medical Imaging, IPMI 2019

Y2 - 2 June 2019 through 7 June 2019

ER -

Deep Modeling of Growth Trajectories for Longitudinal Prediction of Missing Infant Cortical Surfaces

Abstract

Publication series

Conference

Bibliographical note

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this