Deep morphological simplification network (MS-Net) for guided registration of brain magnetic resonance images

Dongming Wei, Lichi Zhang, Zhengwang Wu, Xiaohuan Cao, Gang Li, Dinggang Shen, Qian Wang

Research output: Contribution to journalArticle

Abstract

Deformable brain MR image registration is challenging due to large inter-subject anatomical variation. For example, the highly complex cortical folding pattern makes it hard to accurately align corresponding cortical structures of individual images. In this paper, we propose a novel deep learning way to simplify the difficult registration problem of brain MR images. Specifically, we train a morphological simplification network (MS-Net), which can generate a simple image with less anatomical details based on the complex input. With MS-Net, the complexity of the fixed image or the moving image under registration can be reduced gradually, thus building an individual (simplification) trajectory represented by MS-Net outputs. Since the generated images at the ends of the two trajectories (of the fixed and moving images) are so simple and very similar in appearance, they are easy to register. Thus, the two trajectories can act as a bridge to link the fixed and the moving images, and guide their registration. Our experiments show that the proposed method can achieve highly accurate registration performance on different datasets (i.e., NIREP, LPBA, IBSR, CUMC, and MGH). Moreover, the method can be also easily transferred across diverse image datasets and obtain superior accuracy on surface alignment. We propose MS-Net as a powerful and flexible tool to simplify brain MR images and their registration. To our knowledge, this is the first work to simplify brain MR image registration by deep learning, instead of estimating deformation field directly.

Original languageEnglish
Article number107171
JournalPattern Recognition
Volume100
DOIs
Publication statusPublished - 2020 Apr

Fingerprint

Magnetic resonance
Brain
Image registration
Trajectories
Experiments
Deep learning

Keywords

  • Anatomical complexity
  • Deep learning
  • Deformable image registration

ASJC Scopus subject areas

  • Software
  • Signal Processing
  • Computer Vision and Pattern Recognition
  • Artificial Intelligence

Cite this

Deep morphological simplification network (MS-Net) for guided registration of brain magnetic resonance images. / Wei, Dongming; Zhang, Lichi; Wu, Zhengwang; Cao, Xiaohuan; Li, Gang; Shen, Dinggang; Wang, Qian.

In: Pattern Recognition, Vol. 100, 107171, 04.2020.

Research output: Contribution to journalArticle

Wei, Dongming ; Zhang, Lichi ; Wu, Zhengwang ; Cao, Xiaohuan ; Li, Gang ; Shen, Dinggang ; Wang, Qian. / Deep morphological simplification network (MS-Net) for guided registration of brain magnetic resonance images. In: Pattern Recognition. 2020 ; Vol. 100.
@article{ae96a5d0fc5b4a5287ead6c201e95401,
title = "Deep morphological simplification network (MS-Net) for guided registration of brain magnetic resonance images",
abstract = "Deformable brain MR image registration is challenging due to large inter-subject anatomical variation. For example, the highly complex cortical folding pattern makes it hard to accurately align corresponding cortical structures of individual images. In this paper, we propose a novel deep learning way to simplify the difficult registration problem of brain MR images. Specifically, we train a morphological simplification network (MS-Net), which can generate a simple image with less anatomical details based on the complex input. With MS-Net, the complexity of the fixed image or the moving image under registration can be reduced gradually, thus building an individual (simplification) trajectory represented by MS-Net outputs. Since the generated images at the ends of the two trajectories (of the fixed and moving images) are so simple and very similar in appearance, they are easy to register. Thus, the two trajectories can act as a bridge to link the fixed and the moving images, and guide their registration. Our experiments show that the proposed method can achieve highly accurate registration performance on different datasets (i.e., NIREP, LPBA, IBSR, CUMC, and MGH). Moreover, the method can be also easily transferred across diverse image datasets and obtain superior accuracy on surface alignment. We propose MS-Net as a powerful and flexible tool to simplify brain MR images and their registration. To our knowledge, this is the first work to simplify brain MR image registration by deep learning, instead of estimating deformation field directly.",
keywords = "Anatomical complexity, Deep learning, Deformable image registration",
author = "Dongming Wei and Lichi Zhang and Zhengwang Wu and Xiaohuan Cao and Gang Li and Dinggang Shen and Qian Wang",
year = "2020",
month = "4",
doi = "10.1016/j.patcog.2019.107171",
language = "English",
volume = "100",
journal = "Pattern Recognition",
issn = "0031-3203",
publisher = "Elsevier Limited",

}

TY - JOUR

T1 - Deep morphological simplification network (MS-Net) for guided registration of brain magnetic resonance images

AU - Wei, Dongming

AU - Zhang, Lichi

AU - Wu, Zhengwang

AU - Cao, Xiaohuan

AU - Li, Gang

AU - Shen, Dinggang

AU - Wang, Qian

PY - 2020/4

Y1 - 2020/4

N2 - Deformable brain MR image registration is challenging due to large inter-subject anatomical variation. For example, the highly complex cortical folding pattern makes it hard to accurately align corresponding cortical structures of individual images. In this paper, we propose a novel deep learning way to simplify the difficult registration problem of brain MR images. Specifically, we train a morphological simplification network (MS-Net), which can generate a simple image with less anatomical details based on the complex input. With MS-Net, the complexity of the fixed image or the moving image under registration can be reduced gradually, thus building an individual (simplification) trajectory represented by MS-Net outputs. Since the generated images at the ends of the two trajectories (of the fixed and moving images) are so simple and very similar in appearance, they are easy to register. Thus, the two trajectories can act as a bridge to link the fixed and the moving images, and guide their registration. Our experiments show that the proposed method can achieve highly accurate registration performance on different datasets (i.e., NIREP, LPBA, IBSR, CUMC, and MGH). Moreover, the method can be also easily transferred across diverse image datasets and obtain superior accuracy on surface alignment. We propose MS-Net as a powerful and flexible tool to simplify brain MR images and their registration. To our knowledge, this is the first work to simplify brain MR image registration by deep learning, instead of estimating deformation field directly.

AB - Deformable brain MR image registration is challenging due to large inter-subject anatomical variation. For example, the highly complex cortical folding pattern makes it hard to accurately align corresponding cortical structures of individual images. In this paper, we propose a novel deep learning way to simplify the difficult registration problem of brain MR images. Specifically, we train a morphological simplification network (MS-Net), which can generate a simple image with less anatomical details based on the complex input. With MS-Net, the complexity of the fixed image or the moving image under registration can be reduced gradually, thus building an individual (simplification) trajectory represented by MS-Net outputs. Since the generated images at the ends of the two trajectories (of the fixed and moving images) are so simple and very similar in appearance, they are easy to register. Thus, the two trajectories can act as a bridge to link the fixed and the moving images, and guide their registration. Our experiments show that the proposed method can achieve highly accurate registration performance on different datasets (i.e., NIREP, LPBA, IBSR, CUMC, and MGH). Moreover, the method can be also easily transferred across diverse image datasets and obtain superior accuracy on surface alignment. We propose MS-Net as a powerful and flexible tool to simplify brain MR images and their registration. To our knowledge, this is the first work to simplify brain MR image registration by deep learning, instead of estimating deformation field directly.

KW - Anatomical complexity

KW - Deep learning

KW - Deformable image registration

UR - http://www.scopus.com/inward/record.url?scp=85077344526&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85077344526&partnerID=8YFLogxK

U2 - 10.1016/j.patcog.2019.107171

DO - 10.1016/j.patcog.2019.107171

M3 - Article

AN - SCOPUS:85077344526

VL - 100

JO - Pattern Recognition

JF - Pattern Recognition

SN - 0031-3203

M1 - 107171

ER -