Dual-domain convolutional neural networks for improving structural information in 3 T MRI

Yongqin Zhang, Pew Thian Yap, Liangqiong Qu, Jie Zhi Cheng, Dinggang Shen

Research output: Contribution to journalArticle

1 Citation (Scopus)

Abstract

We propose a novel dual-domain convolutional neural network framework to improve structural information of routine 3 T images. We introduce a parameter-efficient butterfly network that involves two complementary domains: a spatial domain and a frequency domain. The butterfly network allows the interaction of these two domains in learning the complex mapping from 3 T to 7 T images. We verified the efficacy of the dual-domain strategy and butterfly network using 3 T and 7 T image pairs. Experimental results demonstrate that the proposed framework generates synthetic 7 T-like images and achieves performance superior to state-of-the-art methods.

Original languageEnglish
JournalMagnetic Resonance Imaging
DOIs
Publication statusPublished - 2019 Jan 1

Fingerprint

Butterflies
Magnetic resonance imaging
Neural networks
Learning

Keywords

  • Convolutional neural network
  • Deep learning
  • Image super-resolution
  • Image synthesis
  • Magnetic resonance imaging

ASJC Scopus subject areas

  • Biophysics
  • Biomedical Engineering
  • Radiology Nuclear Medicine and imaging

Cite this

Dual-domain convolutional neural networks for improving structural information in 3 T MRI. / Zhang, Yongqin; Yap, Pew Thian; Qu, Liangqiong; Cheng, Jie Zhi; Shen, Dinggang.

In: Magnetic Resonance Imaging, 01.01.2019.

Research output: Contribution to journalArticle

@article{e272c43cebbb4e59807088dea9092c7a,
title = "Dual-domain convolutional neural networks for improving structural information in 3 T MRI",
abstract = "We propose a novel dual-domain convolutional neural network framework to improve structural information of routine 3 T images. We introduce a parameter-efficient butterfly network that involves two complementary domains: a spatial domain and a frequency domain. The butterfly network allows the interaction of these two domains in learning the complex mapping from 3 T to 7 T images. We verified the efficacy of the dual-domain strategy and butterfly network using 3 T and 7 T image pairs. Experimental results demonstrate that the proposed framework generates synthetic 7 T-like images and achieves performance superior to state-of-the-art methods.",
keywords = "Convolutional neural network, Deep learning, Image super-resolution, Image synthesis, Magnetic resonance imaging",
author = "Yongqin Zhang and Yap, {Pew Thian} and Liangqiong Qu and Cheng, {Jie Zhi} and Dinggang Shen",
year = "2019",
month = "1",
day = "1",
doi = "10.1016/j.mri.2019.05.023",
language = "English",
journal = "Magnetic Resonance Imaging",
issn = "0730-725X",
publisher = "Elsevier Inc.",

}

TY - JOUR

T1 - Dual-domain convolutional neural networks for improving structural information in 3 T MRI

AU - Zhang, Yongqin

AU - Yap, Pew Thian

AU - Qu, Liangqiong

AU - Cheng, Jie Zhi

AU - Shen, Dinggang

PY - 2019/1/1

Y1 - 2019/1/1

N2 - We propose a novel dual-domain convolutional neural network framework to improve structural information of routine 3 T images. We introduce a parameter-efficient butterfly network that involves two complementary domains: a spatial domain and a frequency domain. The butterfly network allows the interaction of these two domains in learning the complex mapping from 3 T to 7 T images. We verified the efficacy of the dual-domain strategy and butterfly network using 3 T and 7 T image pairs. Experimental results demonstrate that the proposed framework generates synthetic 7 T-like images and achieves performance superior to state-of-the-art methods.

AB - We propose a novel dual-domain convolutional neural network framework to improve structural information of routine 3 T images. We introduce a parameter-efficient butterfly network that involves two complementary domains: a spatial domain and a frequency domain. The butterfly network allows the interaction of these two domains in learning the complex mapping from 3 T to 7 T images. We verified the efficacy of the dual-domain strategy and butterfly network using 3 T and 7 T image pairs. Experimental results demonstrate that the proposed framework generates synthetic 7 T-like images and achieves performance superior to state-of-the-art methods.

KW - Convolutional neural network

KW - Deep learning

KW - Image super-resolution

KW - Image synthesis

KW - Magnetic resonance imaging

UR - http://www.scopus.com/inward/record.url?scp=85067286440&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85067286440&partnerID=8YFLogxK

U2 - 10.1016/j.mri.2019.05.023

DO - 10.1016/j.mri.2019.05.023

M3 - Article

AN - SCOPUS:85067286440

JO - Magnetic Resonance Imaging

JF - Magnetic Resonance Imaging

SN - 0730-725X

ER -