Topological correction of infant cortical surfaces using anatomically constrained U-net

Liang Sun, Daoqiang Zhang, Li Wang, Wei Shao, Zengsi Chen, Weili Lin, Dinggang Shen, Gang Li

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

Abstract

Reconstruction of accurate cortical surfaces with minimal topological errors (i.e., handles and holes) from infant brain MR images is important in early brain development studies. However, infant brain MR images usually exhibit extremely low tissue contrast (especially from 3 to 9 months of age) and dynamic imaging appearance patterns. Thus, it is inevitable to have large amounts of topological errors in the infant brain tissue segmentation results, thus leading to inaccurate surface reconstruction. To address these issues, inspired by recent advances in deep learning methods, we propose an anatomically constrained U-Net method for topological correction of infant cortical surfaces. Specifically, in our method, we first extract candidate voxels with potential topological errors, by leveraging a topology-preserving level set method. Then, we propose a U-Net with anatomical constraints to correct those located candidate voxels. Due to the fact that infant cortical surfaces often contain large handles or holes, it is difficult to completely correct all errors using one-shot correction. Therefore, we further gather these two steps into an iterative framework to correct large topological errors gradually. To our knowledge, this is the first work introducing deep learning for infant cortical topological correction. We compare our method with the state-of-the-art method on infant cortical topology and show the superior performance of our method.

Original languageEnglish
Title of host publicationMachine Learning in Medical Imaging - 9th International Workshop, MLMI 2018, Held in Conjunction with MICCAI 2018, Proceedings
EditorsMingxia Liu, Heung-Il Suk, Yinghuan Shi
PublisherSpringer Verlag
Pages125-133
Number of pages9
ISBN (Print)9783030009182
DOIs
Publication statusPublished - 2018 Jan 1
Externally publishedYes
Event9th International Workshop on Machine Learning in Medical Imaging, MLMI 2018 held in conjunction with the 21st International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2018 - Granada, Spain
Duration: 2018 Sep 162018 Sep 16

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume11046 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Other

Other9th International Workshop on Machine Learning in Medical Imaging, MLMI 2018 held in conjunction with the 21st International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2018
CountrySpain
CityGranada
Period18/9/1618/9/16

Fingerprint

Brain
Voxel
Topology
Tissue
Surface reconstruction
Surface Reconstruction
Level Set Method
Inaccurate
Imaging techniques
Segmentation
Imaging
Deep learning
Learning

ASJC Scopus subject areas

  • Theoretical Computer Science
  • Computer Science(all)

Cite this

Sun, L., Zhang, D., Wang, L., Shao, W., Chen, Z., Lin, W., ... Li, G. (2018). Topological correction of infant cortical surfaces using anatomically constrained U-net. In M. Liu, H-I. Suk, & Y. Shi (Eds.), Machine Learning in Medical Imaging - 9th International Workshop, MLMI 2018, Held in Conjunction with MICCAI 2018, Proceedings (pp. 125-133). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 11046 LNCS). Springer Verlag. https://doi.org/10.1007/978-3-030-00919-9_15

Topological correction of infant cortical surfaces using anatomically constrained U-net. / Sun, Liang; Zhang, Daoqiang; Wang, Li; Shao, Wei; Chen, Zengsi; Lin, Weili; Shen, Dinggang; Li, Gang.

Machine Learning in Medical Imaging - 9th International Workshop, MLMI 2018, Held in Conjunction with MICCAI 2018, Proceedings. ed. / Mingxia Liu; Heung-Il Suk; Yinghuan Shi. Springer Verlag, 2018. p. 125-133 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 11046 LNCS).

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

Sun, L, Zhang, D, Wang, L, Shao, W, Chen, Z, Lin, W, Shen, D & Li, G 2018, Topological correction of infant cortical surfaces using anatomically constrained U-net. in M Liu, H-I Suk & Y Shi (eds), Machine Learning in Medical Imaging - 9th International Workshop, MLMI 2018, Held in Conjunction with MICCAI 2018, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 11046 LNCS, Springer Verlag, pp. 125-133, 9th International Workshop on Machine Learning in Medical Imaging, MLMI 2018 held in conjunction with the 21st International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2018, Granada, Spain, 18/9/16. https://doi.org/10.1007/978-3-030-00919-9_15
Sun L, Zhang D, Wang L, Shao W, Chen Z, Lin W et al. Topological correction of infant cortical surfaces using anatomically constrained U-net. In Liu M, Suk H-I, Shi Y, editors, Machine Learning in Medical Imaging - 9th International Workshop, MLMI 2018, Held in Conjunction with MICCAI 2018, Proceedings. Springer Verlag. 2018. p. 125-133. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). https://doi.org/10.1007/978-3-030-00919-9_15
Sun, Liang ; Zhang, Daoqiang ; Wang, Li ; Shao, Wei ; Chen, Zengsi ; Lin, Weili ; Shen, Dinggang ; Li, Gang. / Topological correction of infant cortical surfaces using anatomically constrained U-net. Machine Learning in Medical Imaging - 9th International Workshop, MLMI 2018, Held in Conjunction with MICCAI 2018, Proceedings. editor / Mingxia Liu ; Heung-Il Suk ; Yinghuan Shi. Springer Verlag, 2018. pp. 125-133 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
@inproceedings{efecedf0b35448c4bd8fc1561097e94d,
title = "Topological correction of infant cortical surfaces using anatomically constrained U-net",
abstract = "Reconstruction of accurate cortical surfaces with minimal topological errors (i.e., handles and holes) from infant brain MR images is important in early brain development studies. However, infant brain MR images usually exhibit extremely low tissue contrast (especially from 3 to 9 months of age) and dynamic imaging appearance patterns. Thus, it is inevitable to have large amounts of topological errors in the infant brain tissue segmentation results, thus leading to inaccurate surface reconstruction. To address these issues, inspired by recent advances in deep learning methods, we propose an anatomically constrained U-Net method for topological correction of infant cortical surfaces. Specifically, in our method, we first extract candidate voxels with potential topological errors, by leveraging a topology-preserving level set method. Then, we propose a U-Net with anatomical constraints to correct those located candidate voxels. Due to the fact that infant cortical surfaces often contain large handles or holes, it is difficult to completely correct all errors using one-shot correction. Therefore, we further gather these two steps into an iterative framework to correct large topological errors gradually. To our knowledge, this is the first work introducing deep learning for infant cortical topological correction. We compare our method with the state-of-the-art method on infant cortical topology and show the superior performance of our method.",
author = "Liang Sun and Daoqiang Zhang and Li Wang and Wei Shao and Zengsi Chen and Weili Lin and Dinggang Shen and Gang Li",
year = "2018",
month = "1",
day = "1",
doi = "10.1007/978-3-030-00919-9_15",
language = "English",
isbn = "9783030009182",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
publisher = "Springer Verlag",
pages = "125--133",
editor = "Mingxia Liu and Heung-Il Suk and Yinghuan Shi",
booktitle = "Machine Learning in Medical Imaging - 9th International Workshop, MLMI 2018, Held in Conjunction with MICCAI 2018, Proceedings",

}

TY - GEN

T1 - Topological correction of infant cortical surfaces using anatomically constrained U-net

AU - Sun, Liang

AU - Zhang, Daoqiang

AU - Wang, Li

AU - Shao, Wei

AU - Chen, Zengsi

AU - Lin, Weili

AU - Shen, Dinggang

AU - Li, Gang

PY - 2018/1/1

Y1 - 2018/1/1

N2 - Reconstruction of accurate cortical surfaces with minimal topological errors (i.e., handles and holes) from infant brain MR images is important in early brain development studies. However, infant brain MR images usually exhibit extremely low tissue contrast (especially from 3 to 9 months of age) and dynamic imaging appearance patterns. Thus, it is inevitable to have large amounts of topological errors in the infant brain tissue segmentation results, thus leading to inaccurate surface reconstruction. To address these issues, inspired by recent advances in deep learning methods, we propose an anatomically constrained U-Net method for topological correction of infant cortical surfaces. Specifically, in our method, we first extract candidate voxels with potential topological errors, by leveraging a topology-preserving level set method. Then, we propose a U-Net with anatomical constraints to correct those located candidate voxels. Due to the fact that infant cortical surfaces often contain large handles or holes, it is difficult to completely correct all errors using one-shot correction. Therefore, we further gather these two steps into an iterative framework to correct large topological errors gradually. To our knowledge, this is the first work introducing deep learning for infant cortical topological correction. We compare our method with the state-of-the-art method on infant cortical topology and show the superior performance of our method.

AB - Reconstruction of accurate cortical surfaces with minimal topological errors (i.e., handles and holes) from infant brain MR images is important in early brain development studies. However, infant brain MR images usually exhibit extremely low tissue contrast (especially from 3 to 9 months of age) and dynamic imaging appearance patterns. Thus, it is inevitable to have large amounts of topological errors in the infant brain tissue segmentation results, thus leading to inaccurate surface reconstruction. To address these issues, inspired by recent advances in deep learning methods, we propose an anatomically constrained U-Net method for topological correction of infant cortical surfaces. Specifically, in our method, we first extract candidate voxels with potential topological errors, by leveraging a topology-preserving level set method. Then, we propose a U-Net with anatomical constraints to correct those located candidate voxels. Due to the fact that infant cortical surfaces often contain large handles or holes, it is difficult to completely correct all errors using one-shot correction. Therefore, we further gather these two steps into an iterative framework to correct large topological errors gradually. To our knowledge, this is the first work introducing deep learning for infant cortical topological correction. We compare our method with the state-of-the-art method on infant cortical topology and show the superior performance of our method.

UR - http://www.scopus.com/inward/record.url?scp=85054531463&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85054531463&partnerID=8YFLogxK

U2 - 10.1007/978-3-030-00919-9_15

DO - 10.1007/978-3-030-00919-9_15

M3 - Conference contribution

AN - SCOPUS:85054531463

SN - 9783030009182

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

SP - 125

EP - 133

BT - Machine Learning in Medical Imaging - 9th International Workshop, MLMI 2018, Held in Conjunction with MICCAI 2018, Proceedings

A2 - Liu, Mingxia

A2 - Suk, Heung-Il

A2 - Shi, Yinghuan

PB - Springer Verlag

ER -