Multi-atlas based segmentation of brainstem nuclei from MR images by deep hyper-graph learning

Pei Dong, Yangrong Guo, Yue Gao, Peipeng Liang, Yonghong Shi, Qian Wang, Dinggang Shen, Guorong Wu

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

Abstract

Accurate segmentation of brainstem nuclei (red nucleus and substantia nigra) is very important in various neuroimaging applications such as deep brain stimulation and the investigation of imaging biomarkers for Parkinson’s disease (PD). Due to iron deposition during aging, image contrast in the brainstem is very low in Magnetic Resonance (MR) images. Hence, the ambiguity of patch-wise similarity makes the recently successful multi-atlas patch-based label fusion methods have difficulty to perform as competitive as segmenting cortical and sub-cortical regions from MR images. To address this challenge, we propose a novel multi-atlas brainstem nuclei segmentation method using deep hyper-graph learning. Specifically, we achieve this goal in three-fold. First, we employ hyper-graph to combine the advantage of maintaining spatial coherence from graph-based segmentation approaches and the benefit of harnessing population priors from multi-atlas based framework. Second, besides using low-level image appearance, we also extract high-level context features to measure the complex patch-wise relationship. Since the context features are calculated on a tentatively estimated label probability map, we eventually turn our hyper-graph learning based label propagation into a deep and self-refining model. Third, since anatomical labels on some voxels (usually located in uniform regions) can be identified much more reliably than other voxels (usually located at the boundary between two regions), we allow these reliable voxels to propagate their labels to the nearby difficult-to-label voxels. Such hierarchical strategy makes our proposed label fusion method deep and dynamic. We evaluate our proposed label fusion method in segmenting substantia nigra (SN) and red nucleus (RN) from 3.0 T MR images, where our proposed method achieves significant improvement over the state-of-the-art label fusion methods.

Original languageEnglish
Title of host publicationPatch-Based Techniques in Medical Imaging - 2nd International Workshop, Patch-MI 2016 held in conjunction with MICCAI 2016, Proceedings
PublisherSpringer Verlag
Pages51-59
Number of pages9
Volume9993 LNCS
ISBN (Print)9783319471174
DOIs
Publication statusPublished - 2016
Externally publishedYes
Event2nd International Workshop on Patch-Based Techniques in Medical Imaging, Patch-MI 2016 held in conjunction with 19th International Conference on Medical Image Computing and Computer Assisted Intervention, MICCAI 2016 - Athens, Greece
Duration: 2016 Oct 172016 Oct 17

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume9993 LNCS
ISSN (Print)03029743
ISSN (Electronic)16113349

Other

Other2nd International Workshop on Patch-Based Techniques in Medical Imaging, Patch-MI 2016 held in conjunction with 19th International Conference on Medical Image Computing and Computer Assisted Intervention, MICCAI 2016
CountryGreece
CityAthens
Period16/10/1716/10/17

Fingerprint

Magnetic Resonance Image
Atlas
Magnetic resonance
Hypergraph
Nucleus
Labels
Segmentation
Voxel
Fusion
Patch
Fusion reactions
Neuroimaging
Parkinson's Disease
Biomarkers
Threefolds
Iron
Learning
Imaging
Propagation
Refining

ASJC Scopus subject areas

  • Theoretical Computer Science
  • Computer Science(all)

Cite this

Dong, P., Guo, Y., Gao, Y., Liang, P., Shi, Y., Wang, Q., ... Wu, G. (2016). Multi-atlas based segmentation of brainstem nuclei from MR images by deep hyper-graph learning. In Patch-Based Techniques in Medical Imaging - 2nd International Workshop, Patch-MI 2016 held in conjunction with MICCAI 2016, Proceedings (Vol. 9993 LNCS, pp. 51-59). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 9993 LNCS). Springer Verlag. https://doi.org/10.1007/978-3-319-47118-1_7

Multi-atlas based segmentation of brainstem nuclei from MR images by deep hyper-graph learning. / Dong, Pei; Guo, Yangrong; Gao, Yue; Liang, Peipeng; Shi, Yonghong; Wang, Qian; Shen, Dinggang; Wu, Guorong.

Patch-Based Techniques in Medical Imaging - 2nd International Workshop, Patch-MI 2016 held in conjunction with MICCAI 2016, Proceedings. Vol. 9993 LNCS Springer Verlag, 2016. p. 51-59 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 9993 LNCS).

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

Dong, P, Guo, Y, Gao, Y, Liang, P, Shi, Y, Wang, Q, Shen, D & Wu, G 2016, Multi-atlas based segmentation of brainstem nuclei from MR images by deep hyper-graph learning. in Patch-Based Techniques in Medical Imaging - 2nd International Workshop, Patch-MI 2016 held in conjunction with MICCAI 2016, Proceedings. vol. 9993 LNCS, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 9993 LNCS, Springer Verlag, pp. 51-59, 2nd International Workshop on Patch-Based Techniques in Medical Imaging, Patch-MI 2016 held in conjunction with 19th International Conference on Medical Image Computing and Computer Assisted Intervention, MICCAI 2016, Athens, Greece, 16/10/17. https://doi.org/10.1007/978-3-319-47118-1_7
Dong P, Guo Y, Gao Y, Liang P, Shi Y, Wang Q et al. Multi-atlas based segmentation of brainstem nuclei from MR images by deep hyper-graph learning. In Patch-Based Techniques in Medical Imaging - 2nd International Workshop, Patch-MI 2016 held in conjunction with MICCAI 2016, Proceedings. Vol. 9993 LNCS. Springer Verlag. 2016. p. 51-59. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). https://doi.org/10.1007/978-3-319-47118-1_7
Dong, Pei ; Guo, Yangrong ; Gao, Yue ; Liang, Peipeng ; Shi, Yonghong ; Wang, Qian ; Shen, Dinggang ; Wu, Guorong. / Multi-atlas based segmentation of brainstem nuclei from MR images by deep hyper-graph learning. Patch-Based Techniques in Medical Imaging - 2nd International Workshop, Patch-MI 2016 held in conjunction with MICCAI 2016, Proceedings. Vol. 9993 LNCS Springer Verlag, 2016. pp. 51-59 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
@inproceedings{55c2249cda6d4b2990de7f21143e6c2b,
title = "Multi-atlas based segmentation of brainstem nuclei from MR images by deep hyper-graph learning",
abstract = "Accurate segmentation of brainstem nuclei (red nucleus and substantia nigra) is very important in various neuroimaging applications such as deep brain stimulation and the investigation of imaging biomarkers for Parkinson’s disease (PD). Due to iron deposition during aging, image contrast in the brainstem is very low in Magnetic Resonance (MR) images. Hence, the ambiguity of patch-wise similarity makes the recently successful multi-atlas patch-based label fusion methods have difficulty to perform as competitive as segmenting cortical and sub-cortical regions from MR images. To address this challenge, we propose a novel multi-atlas brainstem nuclei segmentation method using deep hyper-graph learning. Specifically, we achieve this goal in three-fold. First, we employ hyper-graph to combine the advantage of maintaining spatial coherence from graph-based segmentation approaches and the benefit of harnessing population priors from multi-atlas based framework. Second, besides using low-level image appearance, we also extract high-level context features to measure the complex patch-wise relationship. Since the context features are calculated on a tentatively estimated label probability map, we eventually turn our hyper-graph learning based label propagation into a deep and self-refining model. Third, since anatomical labels on some voxels (usually located in uniform regions) can be identified much more reliably than other voxels (usually located at the boundary between two regions), we allow these reliable voxels to propagate their labels to the nearby difficult-to-label voxels. Such hierarchical strategy makes our proposed label fusion method deep and dynamic. We evaluate our proposed label fusion method in segmenting substantia nigra (SN) and red nucleus (RN) from 3.0 T MR images, where our proposed method achieves significant improvement over the state-of-the-art label fusion methods.",
author = "Pei Dong and Yangrong Guo and Yue Gao and Peipeng Liang and Yonghong Shi and Qian Wang and Dinggang Shen and Guorong Wu",
year = "2016",
doi = "10.1007/978-3-319-47118-1_7",
language = "English",
isbn = "9783319471174",
volume = "9993 LNCS",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
publisher = "Springer Verlag",
pages = "51--59",
booktitle = "Patch-Based Techniques in Medical Imaging - 2nd International Workshop, Patch-MI 2016 held in conjunction with MICCAI 2016, Proceedings",

}

TY - GEN

T1 - Multi-atlas based segmentation of brainstem nuclei from MR images by deep hyper-graph learning

AU - Dong, Pei

AU - Guo, Yangrong

AU - Gao, Yue

AU - Liang, Peipeng

AU - Shi, Yonghong

AU - Wang, Qian

AU - Shen, Dinggang

AU - Wu, Guorong

PY - 2016

Y1 - 2016

N2 - Accurate segmentation of brainstem nuclei (red nucleus and substantia nigra) is very important in various neuroimaging applications such as deep brain stimulation and the investigation of imaging biomarkers for Parkinson’s disease (PD). Due to iron deposition during aging, image contrast in the brainstem is very low in Magnetic Resonance (MR) images. Hence, the ambiguity of patch-wise similarity makes the recently successful multi-atlas patch-based label fusion methods have difficulty to perform as competitive as segmenting cortical and sub-cortical regions from MR images. To address this challenge, we propose a novel multi-atlas brainstem nuclei segmentation method using deep hyper-graph learning. Specifically, we achieve this goal in three-fold. First, we employ hyper-graph to combine the advantage of maintaining spatial coherence from graph-based segmentation approaches and the benefit of harnessing population priors from multi-atlas based framework. Second, besides using low-level image appearance, we also extract high-level context features to measure the complex patch-wise relationship. Since the context features are calculated on a tentatively estimated label probability map, we eventually turn our hyper-graph learning based label propagation into a deep and self-refining model. Third, since anatomical labels on some voxels (usually located in uniform regions) can be identified much more reliably than other voxels (usually located at the boundary between two regions), we allow these reliable voxels to propagate their labels to the nearby difficult-to-label voxels. Such hierarchical strategy makes our proposed label fusion method deep and dynamic. We evaluate our proposed label fusion method in segmenting substantia nigra (SN) and red nucleus (RN) from 3.0 T MR images, where our proposed method achieves significant improvement over the state-of-the-art label fusion methods.

AB - Accurate segmentation of brainstem nuclei (red nucleus and substantia nigra) is very important in various neuroimaging applications such as deep brain stimulation and the investigation of imaging biomarkers for Parkinson’s disease (PD). Due to iron deposition during aging, image contrast in the brainstem is very low in Magnetic Resonance (MR) images. Hence, the ambiguity of patch-wise similarity makes the recently successful multi-atlas patch-based label fusion methods have difficulty to perform as competitive as segmenting cortical and sub-cortical regions from MR images. To address this challenge, we propose a novel multi-atlas brainstem nuclei segmentation method using deep hyper-graph learning. Specifically, we achieve this goal in three-fold. First, we employ hyper-graph to combine the advantage of maintaining spatial coherence from graph-based segmentation approaches and the benefit of harnessing population priors from multi-atlas based framework. Second, besides using low-level image appearance, we also extract high-level context features to measure the complex patch-wise relationship. Since the context features are calculated on a tentatively estimated label probability map, we eventually turn our hyper-graph learning based label propagation into a deep and self-refining model. Third, since anatomical labels on some voxels (usually located in uniform regions) can be identified much more reliably than other voxels (usually located at the boundary between two regions), we allow these reliable voxels to propagate their labels to the nearby difficult-to-label voxels. Such hierarchical strategy makes our proposed label fusion method deep and dynamic. We evaluate our proposed label fusion method in segmenting substantia nigra (SN) and red nucleus (RN) from 3.0 T MR images, where our proposed method achieves significant improvement over the state-of-the-art label fusion methods.

UR - http://www.scopus.com/inward/record.url?scp=84992504884&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84992504884&partnerID=8YFLogxK

U2 - 10.1007/978-3-319-47118-1_7

DO - 10.1007/978-3-319-47118-1_7

M3 - Conference contribution

AN - SCOPUS:84992504884

SN - 9783319471174

VL - 9993 LNCS

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

SP - 51

EP - 59

BT - Patch-Based Techniques in Medical Imaging - 2nd International Workshop, Patch-MI 2016 held in conjunction with MICCAI 2016, Proceedings

PB - Springer Verlag

ER -