Fiber orientation estimation guided by a deep network

Chuyang Ye, Jerry Ladd Prince

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

Abstract

Diffusion magnetic resonance imaging (dMRI) is currently the only tool for noninvasively imaging the brain’s white matter tracts. The fiber orientation (FO) is a key feature computed from dMRI for tract reconstruction. Because the number of FOs in a voxel is usually small, dictionary-based sparse reconstruction has been used to estimate FOs. However, accurate estimation of complex FO configurations in the presence of noise can still be challenging. In this work we explore the use of a deep network for FO estimation in a dictionary-based framework and propose an algorithm named Fiber Orientation Reconstruction guided by a Deep Network (FORDN). FORDN consists of two steps. First, we use a smaller dictionary encoding coarse basis FOs to represent diffusion signals. To estimate the mixture fractions of the dictionary atoms, a deep network is designed to solve the sparse reconstruction problem. Second, the coarse FOs inform the final FO estimation, where a larger dictionary encoding a dense basis of FOs is used and a weighted ℓ1 -norm regularized least squares problem is solved to encourage FOs that are consistent with the network output. FORDN was evaluated and compared with state-of-the-art algorithms that estimate FOs using sparse reconstruction on simulated and typical clinical dMRI data. The results demonstrate the benefit of using a deep network for FO estimation.

Original languageEnglish (US)
Title of host publicationMedical Image Computing and Computer Assisted Intervention − MICCAI 2017 - 20th International Conference, Proceedings
EditorsMaxime Descoteaux, Simon Duchesne, Alfred Franz, Pierre Jannin, D. Louis Collins, Lena Maier-Hein
PublisherSpringer Verlag
Pages575-583
Number of pages9
ISBN (Print)9783319661810
DOIs
StatePublished - Jan 1 2017
Event20th International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2017 - Quebec City, Canada
Duration: Sep 11 2017Sep 13 2017

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume10433 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Other

Other20th International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2017
CountryCanada
CityQuebec City
Period9/11/179/13/17

Fingerprint

Fiber Orientation
Fiber reinforced materials
Glossaries
Magnetic Resonance Imaging
Magnetic resonance
Imaging techniques
Encoding
Estimate
Least Squares Problem
Voxel
Brain
Imaging
Dictionary
Norm
Atoms
Configuration
Output

Keywords

  • Deep Network
  • Diffusion MRI
  • Fiber orientation estimation
  • Sparse reconstruction

ASJC Scopus subject areas

  • Theoretical Computer Science
  • Computer Science(all)

Cite this

Ye, C., & Prince, J. L. (2017). Fiber orientation estimation guided by a deep network. In M. Descoteaux, S. Duchesne, A. Franz, P. Jannin, D. L. Collins, & L. Maier-Hein (Eds.), Medical Image Computing and Computer Assisted Intervention − MICCAI 2017 - 20th International Conference, Proceedings (pp. 575-583). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 10433 LNCS). Springer Verlag. https://doi.org/10.1007/978-3-319-66182-7_66

Fiber orientation estimation guided by a deep network. / Ye, Chuyang; Prince, Jerry Ladd.

Medical Image Computing and Computer Assisted Intervention − MICCAI 2017 - 20th International Conference, Proceedings. ed. / Maxime Descoteaux; Simon Duchesne; Alfred Franz; Pierre Jannin; D. Louis Collins; Lena Maier-Hein. Springer Verlag, 2017. p. 575-583 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 10433 LNCS).

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

Ye, C & Prince, JL 2017, Fiber orientation estimation guided by a deep network. in M Descoteaux, S Duchesne, A Franz, P Jannin, DL Collins & L Maier-Hein (eds), Medical Image Computing and Computer Assisted Intervention − MICCAI 2017 - 20th International Conference, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 10433 LNCS, Springer Verlag, pp. 575-583, 20th International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2017, Quebec City, Canada, 9/11/17. https://doi.org/10.1007/978-3-319-66182-7_66
Ye C, Prince JL. Fiber orientation estimation guided by a deep network. In Descoteaux M, Duchesne S, Franz A, Jannin P, Collins DL, Maier-Hein L, editors, Medical Image Computing and Computer Assisted Intervention − MICCAI 2017 - 20th International Conference, Proceedings. Springer Verlag. 2017. p. 575-583. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). https://doi.org/10.1007/978-3-319-66182-7_66
Ye, Chuyang ; Prince, Jerry Ladd. / Fiber orientation estimation guided by a deep network. Medical Image Computing and Computer Assisted Intervention − MICCAI 2017 - 20th International Conference, Proceedings. editor / Maxime Descoteaux ; Simon Duchesne ; Alfred Franz ; Pierre Jannin ; D. Louis Collins ; Lena Maier-Hein. Springer Verlag, 2017. pp. 575-583 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
@inproceedings{ddf0d34917f4462c8e74d967fa54f2e2,
title = "Fiber orientation estimation guided by a deep network",
abstract = "Diffusion magnetic resonance imaging (dMRI) is currently the only tool for noninvasively imaging the brain’s white matter tracts. The fiber orientation (FO) is a key feature computed from dMRI for tract reconstruction. Because the number of FOs in a voxel is usually small, dictionary-based sparse reconstruction has been used to estimate FOs. However, accurate estimation of complex FO configurations in the presence of noise can still be challenging. In this work we explore the use of a deep network for FO estimation in a dictionary-based framework and propose an algorithm named Fiber Orientation Reconstruction guided by a Deep Network (FORDN). FORDN consists of two steps. First, we use a smaller dictionary encoding coarse basis FOs to represent diffusion signals. To estimate the mixture fractions of the dictionary atoms, a deep network is designed to solve the sparse reconstruction problem. Second, the coarse FOs inform the final FO estimation, where a larger dictionary encoding a dense basis of FOs is used and a weighted ℓ1 -norm regularized least squares problem is solved to encourage FOs that are consistent with the network output. FORDN was evaluated and compared with state-of-the-art algorithms that estimate FOs using sparse reconstruction on simulated and typical clinical dMRI data. The results demonstrate the benefit of using a deep network for FO estimation.",
keywords = "Deep Network, Diffusion MRI, Fiber orientation estimation, Sparse reconstruction",
author = "Chuyang Ye and Prince, {Jerry Ladd}",
year = "2017",
month = "1",
day = "1",
doi = "10.1007/978-3-319-66182-7_66",
language = "English (US)",
isbn = "9783319661810",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
publisher = "Springer Verlag",
pages = "575--583",
editor = "Maxime Descoteaux and Simon Duchesne and Alfred Franz and Pierre Jannin and Collins, {D. Louis} and Lena Maier-Hein",
booktitle = "Medical Image Computing and Computer Assisted Intervention − MICCAI 2017 - 20th International Conference, Proceedings",

}

TY - GEN

T1 - Fiber orientation estimation guided by a deep network

AU - Ye, Chuyang

AU - Prince, Jerry Ladd

PY - 2017/1/1

Y1 - 2017/1/1

N2 - Diffusion magnetic resonance imaging (dMRI) is currently the only tool for noninvasively imaging the brain’s white matter tracts. The fiber orientation (FO) is a key feature computed from dMRI for tract reconstruction. Because the number of FOs in a voxel is usually small, dictionary-based sparse reconstruction has been used to estimate FOs. However, accurate estimation of complex FO configurations in the presence of noise can still be challenging. In this work we explore the use of a deep network for FO estimation in a dictionary-based framework and propose an algorithm named Fiber Orientation Reconstruction guided by a Deep Network (FORDN). FORDN consists of two steps. First, we use a smaller dictionary encoding coarse basis FOs to represent diffusion signals. To estimate the mixture fractions of the dictionary atoms, a deep network is designed to solve the sparse reconstruction problem. Second, the coarse FOs inform the final FO estimation, where a larger dictionary encoding a dense basis of FOs is used and a weighted ℓ1 -norm regularized least squares problem is solved to encourage FOs that are consistent with the network output. FORDN was evaluated and compared with state-of-the-art algorithms that estimate FOs using sparse reconstruction on simulated and typical clinical dMRI data. The results demonstrate the benefit of using a deep network for FO estimation.

AB - Diffusion magnetic resonance imaging (dMRI) is currently the only tool for noninvasively imaging the brain’s white matter tracts. The fiber orientation (FO) is a key feature computed from dMRI for tract reconstruction. Because the number of FOs in a voxel is usually small, dictionary-based sparse reconstruction has been used to estimate FOs. However, accurate estimation of complex FO configurations in the presence of noise can still be challenging. In this work we explore the use of a deep network for FO estimation in a dictionary-based framework and propose an algorithm named Fiber Orientation Reconstruction guided by a Deep Network (FORDN). FORDN consists of two steps. First, we use a smaller dictionary encoding coarse basis FOs to represent diffusion signals. To estimate the mixture fractions of the dictionary atoms, a deep network is designed to solve the sparse reconstruction problem. Second, the coarse FOs inform the final FO estimation, where a larger dictionary encoding a dense basis of FOs is used and a weighted ℓ1 -norm regularized least squares problem is solved to encourage FOs that are consistent with the network output. FORDN was evaluated and compared with state-of-the-art algorithms that estimate FOs using sparse reconstruction on simulated and typical clinical dMRI data. The results demonstrate the benefit of using a deep network for FO estimation.

KW - Deep Network

KW - Diffusion MRI

KW - Fiber orientation estimation

KW - Sparse reconstruction

UR - http://www.scopus.com/inward/record.url?scp=85029374625&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85029374625&partnerID=8YFLogxK

U2 - 10.1007/978-3-319-66182-7_66

DO - 10.1007/978-3-319-66182-7_66

M3 - Conference contribution

C2 - 28944347

AN - SCOPUS:85029374625

SN - 9783319661810

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

SP - 575

EP - 583

BT - Medical Image Computing and Computer Assisted Intervention − MICCAI 2017 - 20th International Conference, Proceedings

A2 - Descoteaux, Maxime

A2 - Duchesne, Simon

A2 - Franz, Alfred

A2 - Jannin, Pierre

A2 - Collins, D. Louis

A2 - Maier-Hein, Lena

PB - Springer Verlag

ER -