Benchmark on automatic 6-month-old infant brain segmentation algorithms: The iSeg-2017 Challenge

Li Wang (Corresponding author), Dong Nie, Guannan Li, Elodie Puybareau, Jose Dolz, Qian Zhang, Fan Wang, Jing Xia, Zhengwang Wu, Jiawei Chen, Kim-Han Thung, Toan Duc Bui, Jitae Shin, Guodong Zeng, Guoyan Zheng, Vladimir S. Fonov, Andrew Doyle, Yongchao Xu, Pim Moeskops, Josien P.W. Pluim & 9 others Christian Desrosiers, Ismail Ben Ayed, Gerard Sanroma, Oualid M. Benkarim, Adria Casamitjana, Veronica Vilaplana, Weili Lin, Gang Li, Dinggang Shen

Research output: Contribution to journalArticleAcademicpeer-review

Abstract

Accurate segmentation of infant brain magnetic resonance (MR) images into white matter (WM), gray matter (GM), and cerebrospinal fluid (CSF) is an indispensable foundation for early studying of brain growth patterns and morphological changes in neurodevelopmental disorders. Nevertheless, in the isointense phase (approximately 6-9 months of age), due to inherent myelination and maturation process, WM and GM exhibit similar levels of intensity in both T1-weighted (T1w) and T2-weighted (T2w) MR images, making tissue segmentation very challenging. Despite many efforts were devoted to brain segmentation, only few studies have focused on the segmentation of 6-month infant brain images. With the idea of boosting methodological development in the community, iSeg-2017 challenge (http://iseg2017.web.unc.edu) provides a set of 6-month infant subjects with manual labels for training and testing the participating methods. Among the 21 automatic segmentation methods participating in iSeg-2017, we review the 8 top-ranked teams, in terms of Dice ratio, modified Hausdorff distance and average surface distance, and introduce their pipelines, implementations, as well as source codes. We further discuss limitations and possible future directions. We hope the dataset in iSeg-2017 and this review article could provide insights into methodological development for the community.

LanguageEnglish
Pages2219-2230
JournalIEEE Transactions on Medical Imaging
Volume38
Issue number9
DOIs
StatePublished - 27 Feb 2019

Fingerprint

Benchmarking
Brain
Magnetic resonance
Magnetic Resonance Spectroscopy
Cerebrospinal fluid
Cerebrospinal Fluid
Labels
Pipelines
Tissue
Testing
Growth
White Matter
Gray Matter

Cite this

Wang, L., Nie, D., Li, G., Puybareau, E., Dolz, J., Zhang, Q., ... Shen, D. (2019). Benchmark on automatic 6-month-old infant brain segmentation algorithms: The iSeg-2017 Challenge. IEEE Transactions on Medical Imaging, 38(9), 2219-2230. DOI: 10.1109/TMI.2019.2901712
Wang, Li ; Nie, Dong ; Li, Guannan ; Puybareau, Elodie ; Dolz, Jose ; Zhang, Qian ; Wang, Fan ; Xia, Jing ; Wu, Zhengwang ; Chen, Jiawei ; Thung, Kim-Han ; Bui, Toan Duc ; Shin, Jitae ; Zeng, Guodong ; Zheng, Guoyan ; Fonov, Vladimir S. ; Doyle, Andrew ; Xu, Yongchao ; Moeskops, Pim ; Pluim, Josien P.W. ; Desrosiers, Christian ; Ayed, Ismail Ben ; Sanroma, Gerard ; Benkarim, Oualid M. ; Casamitjana, Adria ; Vilaplana, Veronica ; Lin, Weili ; Li, Gang ; Shen, Dinggang. / Benchmark on automatic 6-month-old infant brain segmentation algorithms : The iSeg-2017 Challenge. In: IEEE Transactions on Medical Imaging. 2019 ; Vol. 38, No. 9. pp. 2219-2230
@article{cf58fbb7e54c41b2a5ab06842c220a32,
title = "Benchmark on automatic 6-month-old infant brain segmentation algorithms: The iSeg-2017 Challenge",
abstract = "Accurate segmentation of infant brain magnetic resonance (MR) images into white matter (WM), gray matter (GM), and cerebrospinal fluid (CSF) is an indispensable foundation for early studying of brain growth patterns and morphological changes in neurodevelopmental disorders. Nevertheless, in the isointense phase (approximately 6-9 months of age), due to inherent myelination and maturation process, WM and GM exhibit similar levels of intensity in both T1-weighted (T1w) and T2-weighted (T2w) MR images, making tissue segmentation very challenging. Despite many efforts were devoted to brain segmentation, only few studies have focused on the segmentation of 6-month infant brain images. With the idea of boosting methodological development in the community, iSeg-2017 challenge (http://iseg2017.web.unc.edu) provides a set of 6-month infant subjects with manual labels for training and testing the participating methods. Among the 21 automatic segmentation methods participating in iSeg-2017, we review the 8 top-ranked teams, in terms of Dice ratio, modified Hausdorff distance and average surface distance, and introduce their pipelines, implementations, as well as source codes. We further discuss limitations and possible future directions. We hope the dataset in iSeg-2017 and this review article could provide insights into methodological development for the community.",
author = "Li Wang and Dong Nie and Guannan Li and Elodie Puybareau and Jose Dolz and Qian Zhang and Fan Wang and Jing Xia and Zhengwang Wu and Jiawei Chen and Kim-Han Thung and Bui, {Toan Duc} and Jitae Shin and Guodong Zeng and Guoyan Zheng and Fonov, {Vladimir S.} and Andrew Doyle and Yongchao Xu and Pim Moeskops and Pluim, {Josien P.W.} and Christian Desrosiers and Ayed, {Ismail Ben} and Gerard Sanroma and Benkarim, {Oualid M.} and Adria Casamitjana and Veronica Vilaplana and Weili Lin and Gang Li and Dinggang Shen",
year = "2019",
month = "2",
day = "27",
doi = "10.1109/TMI.2019.2901712",
language = "English",
volume = "38",
pages = "2219--2230",
journal = "IEEE Transactions on Medical Imaging",
issn = "0278-0062",
publisher = "Institute of Electrical and Electronics Engineers",
number = "9",

}

Wang, L, Nie, D, Li, G, Puybareau, E, Dolz, J, Zhang, Q, Wang, F, Xia, J, Wu, Z, Chen, J, Thung, K-H, Bui, TD, Shin, J, Zeng, G, Zheng, G, Fonov, VS, Doyle, A, Xu, Y, Moeskops, P, Pluim, JPW, Desrosiers, C, Ayed, IB, Sanroma, G, Benkarim, OM, Casamitjana, A, Vilaplana, V, Lin, W, Li, G & Shen, D 2019, 'Benchmark on automatic 6-month-old infant brain segmentation algorithms: The iSeg-2017 Challenge' IEEE Transactions on Medical Imaging, vol. 38, no. 9, pp. 2219-2230. DOI: 10.1109/TMI.2019.2901712

Benchmark on automatic 6-month-old infant brain segmentation algorithms : The iSeg-2017 Challenge. / Wang, Li (Corresponding author); Nie, Dong; Li, Guannan; Puybareau, Elodie; Dolz, Jose; Zhang, Qian; Wang, Fan; Xia, Jing; Wu, Zhengwang; Chen, Jiawei; Thung, Kim-Han; Bui, Toan Duc; Shin, Jitae; Zeng, Guodong; Zheng, Guoyan; Fonov, Vladimir S.; Doyle, Andrew; Xu, Yongchao; Moeskops, Pim; Pluim, Josien P.W.; Desrosiers, Christian; Ayed, Ismail Ben; Sanroma, Gerard; Benkarim, Oualid M.; Casamitjana, Adria; Vilaplana, Veronica; Lin, Weili; Li, Gang; Shen, Dinggang (Corresponding author).

In: IEEE Transactions on Medical Imaging, Vol. 38, No. 9, 27.02.2019, p. 2219-2230.

Research output: Contribution to journalArticleAcademicpeer-review

TY - JOUR

T1 - Benchmark on automatic 6-month-old infant brain segmentation algorithms

T2 - IEEE Transactions on Medical Imaging

AU - Wang,Li

AU - Nie,Dong

AU - Li,Guannan

AU - Puybareau,Elodie

AU - Dolz,Jose

AU - Zhang,Qian

AU - Wang,Fan

AU - Xia,Jing

AU - Wu,Zhengwang

AU - Chen,Jiawei

AU - Thung,Kim-Han

AU - Bui,Toan Duc

AU - Shin,Jitae

AU - Zeng,Guodong

AU - Zheng,Guoyan

AU - Fonov,Vladimir S.

AU - Doyle,Andrew

AU - Xu,Yongchao

AU - Moeskops,Pim

AU - Pluim,Josien P.W.

AU - Desrosiers,Christian

AU - Ayed,Ismail Ben

AU - Sanroma,Gerard

AU - Benkarim,Oualid M.

AU - Casamitjana,Adria

AU - Vilaplana,Veronica

AU - Lin,Weili

AU - Li,Gang

AU - Shen,Dinggang

PY - 2019/2/27

Y1 - 2019/2/27

N2 - Accurate segmentation of infant brain magnetic resonance (MR) images into white matter (WM), gray matter (GM), and cerebrospinal fluid (CSF) is an indispensable foundation for early studying of brain growth patterns and morphological changes in neurodevelopmental disorders. Nevertheless, in the isointense phase (approximately 6-9 months of age), due to inherent myelination and maturation process, WM and GM exhibit similar levels of intensity in both T1-weighted (T1w) and T2-weighted (T2w) MR images, making tissue segmentation very challenging. Despite many efforts were devoted to brain segmentation, only few studies have focused on the segmentation of 6-month infant brain images. With the idea of boosting methodological development in the community, iSeg-2017 challenge (http://iseg2017.web.unc.edu) provides a set of 6-month infant subjects with manual labels for training and testing the participating methods. Among the 21 automatic segmentation methods participating in iSeg-2017, we review the 8 top-ranked teams, in terms of Dice ratio, modified Hausdorff distance and average surface distance, and introduce their pipelines, implementations, as well as source codes. We further discuss limitations and possible future directions. We hope the dataset in iSeg-2017 and this review article could provide insights into methodological development for the community.

AB - Accurate segmentation of infant brain magnetic resonance (MR) images into white matter (WM), gray matter (GM), and cerebrospinal fluid (CSF) is an indispensable foundation for early studying of brain growth patterns and morphological changes in neurodevelopmental disorders. Nevertheless, in the isointense phase (approximately 6-9 months of age), due to inherent myelination and maturation process, WM and GM exhibit similar levels of intensity in both T1-weighted (T1w) and T2-weighted (T2w) MR images, making tissue segmentation very challenging. Despite many efforts were devoted to brain segmentation, only few studies have focused on the segmentation of 6-month infant brain images. With the idea of boosting methodological development in the community, iSeg-2017 challenge (http://iseg2017.web.unc.edu) provides a set of 6-month infant subjects with manual labels for training and testing the participating methods. Among the 21 automatic segmentation methods participating in iSeg-2017, we review the 8 top-ranked teams, in terms of Dice ratio, modified Hausdorff distance and average surface distance, and introduce their pipelines, implementations, as well as source codes. We further discuss limitations and possible future directions. We hope the dataset in iSeg-2017 and this review article could provide insights into methodological development for the community.

U2 - 10.1109/TMI.2019.2901712

DO - 10.1109/TMI.2019.2901712

M3 - Article

VL - 38

SP - 2219

EP - 2230

JO - IEEE Transactions on Medical Imaging

JF - IEEE Transactions on Medical Imaging

SN - 0278-0062

IS - 9

ER -

Wang L, Nie D, Li G, Puybareau E, Dolz J, Zhang Q et al. Benchmark on automatic 6-month-old infant brain segmentation algorithms: The iSeg-2017 Challenge. IEEE Transactions on Medical Imaging. 2019 Feb 27;38(9):2219-2230. Available from, DOI: 10.1109/TMI.2019.2901712