Longitudinal Prediction of Infant Diffusion MRI Data via Graph Convolutional Adversarial Networks

Yoonmi Hong, Jaeil Kim, Geng Chen, Weili Lin, Pew Thian Yap, Dinggang Shen

Research output: Contribution to journalArticle

2 Citations (Scopus)

Abstract

Missing data is a common problem in longitudinal studies due to subject dropouts and failed scans. We present a graph-based convolutional neural network to predict missing diffusion MRI data. In particular, we consider the relationships between sampling points in the spatial domain and the diffusion wave-vector domain to construct a graph. We then use a graph convolutional network to learn the non-linear mapping from available data to missing data. Our method harnesses a multi-scale residual architecture with adversarial learning for prediction with greater accuracy and perceptual quality. Experimental results show that our method is accurate and robust in the longitudinal prediction of infant brain diffusion MRI data.

Original languageEnglish
Article number8691605
Pages (from-to)2717-2725
Number of pages9
JournalIEEE Transactions on Medical Imaging
Volume38
Issue number12
DOIs
Publication statusPublished - 2019 Dec

Fingerprint

Diffusion Magnetic Resonance Imaging
Magnetic resonance imaging
Longitudinal Studies
Learning
Brain
Sampling
Neural networks

Keywords

  • adversarial learning
  • diffusion MRI
  • early brain development
  • Graph CNN
  • longitudinal prediction

ASJC Scopus subject areas

  • Software
  • Radiological and Ultrasound Technology
  • Computer Science Applications
  • Electrical and Electronic Engineering

Cite this

Longitudinal Prediction of Infant Diffusion MRI Data via Graph Convolutional Adversarial Networks. / Hong, Yoonmi; Kim, Jaeil; Chen, Geng; Lin, Weili; Yap, Pew Thian; Shen, Dinggang.

In: IEEE Transactions on Medical Imaging, Vol. 38, No. 12, 8691605, 12.2019, p. 2717-2725.

Research output: Contribution to journalArticle

Hong, Yoonmi ; Kim, Jaeil ; Chen, Geng ; Lin, Weili ; Yap, Pew Thian ; Shen, Dinggang. / Longitudinal Prediction of Infant Diffusion MRI Data via Graph Convolutional Adversarial Networks. In: IEEE Transactions on Medical Imaging. 2019 ; Vol. 38, No. 12. pp. 2717-2725.
@article{6724f68cd4b946cf9013d9f96e177e77,
title = "Longitudinal Prediction of Infant Diffusion MRI Data via Graph Convolutional Adversarial Networks",
abstract = "Missing data is a common problem in longitudinal studies due to subject dropouts and failed scans. We present a graph-based convolutional neural network to predict missing diffusion MRI data. In particular, we consider the relationships between sampling points in the spatial domain and the diffusion wave-vector domain to construct a graph. We then use a graph convolutional network to learn the non-linear mapping from available data to missing data. Our method harnesses a multi-scale residual architecture with adversarial learning for prediction with greater accuracy and perceptual quality. Experimental results show that our method is accurate and robust in the longitudinal prediction of infant brain diffusion MRI data.",
keywords = "adversarial learning, diffusion MRI, early brain development, Graph CNN, longitudinal prediction",
author = "Yoonmi Hong and Jaeil Kim and Geng Chen and Weili Lin and Yap, {Pew Thian} and Dinggang Shen",
year = "2019",
month = "12",
doi = "10.1109/TMI.2019.2911203",
language = "English",
volume = "38",
pages = "2717--2725",
journal = "IEEE Transactions on Medical Imaging",
issn = "0278-0062",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
number = "12",

}

TY - JOUR

T1 - Longitudinal Prediction of Infant Diffusion MRI Data via Graph Convolutional Adversarial Networks

AU - Hong, Yoonmi

AU - Kim, Jaeil

AU - Chen, Geng

AU - Lin, Weili

AU - Yap, Pew Thian

AU - Shen, Dinggang

PY - 2019/12

Y1 - 2019/12

N2 - Missing data is a common problem in longitudinal studies due to subject dropouts and failed scans. We present a graph-based convolutional neural network to predict missing diffusion MRI data. In particular, we consider the relationships between sampling points in the spatial domain and the diffusion wave-vector domain to construct a graph. We then use a graph convolutional network to learn the non-linear mapping from available data to missing data. Our method harnesses a multi-scale residual architecture with adversarial learning for prediction with greater accuracy and perceptual quality. Experimental results show that our method is accurate and robust in the longitudinal prediction of infant brain diffusion MRI data.

AB - Missing data is a common problem in longitudinal studies due to subject dropouts and failed scans. We present a graph-based convolutional neural network to predict missing diffusion MRI data. In particular, we consider the relationships between sampling points in the spatial domain and the diffusion wave-vector domain to construct a graph. We then use a graph convolutional network to learn the non-linear mapping from available data to missing data. Our method harnesses a multi-scale residual architecture with adversarial learning for prediction with greater accuracy and perceptual quality. Experimental results show that our method is accurate and robust in the longitudinal prediction of infant brain diffusion MRI data.

KW - adversarial learning

KW - diffusion MRI

KW - early brain development

KW - Graph CNN

KW - longitudinal prediction

UR - http://www.scopus.com/inward/record.url?scp=85068232902&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85068232902&partnerID=8YFLogxK

U2 - 10.1109/TMI.2019.2911203

DO - 10.1109/TMI.2019.2911203

M3 - Article

C2 - 30990424

AN - SCOPUS:85068232902

VL - 38

SP - 2717

EP - 2725

JO - IEEE Transactions on Medical Imaging

JF - IEEE Transactions on Medical Imaging

SN - 0278-0062

IS - 12

M1 - 8691605

ER -