LATEST: Local AdapTivE and Sequential Training for tissue segmentation of isointense infant brain MR images

Li Wang, Yaozong Gao, Gang Li, Feng Shi, Weili Lin, Dinggang Shen

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

Abstract

Accurate segmentation of isointense infant (~6 months of age) brain MRIs is of great importance, however, a very challenging task, due to extremely low tissue contrast caused by ongoing myelination processes. In this work, we propose a novel learning method based on Local AdapTivE and Sequential Training (LATEST) for segmentation. Specifically, random forest technique is employed to train a local classifier (a single decision tree) for each voxel in the common space based on the neighboring training samples from atlases. Then, for each given voxel, all trained nearby individual classifiers (decision trees) are grouped together to form a forest. Moreover, the estimated probabilities are further used as additional source images to train the next set of local classifiers for refining tissue classification. By iteratively training the subsequent classifiers based on the updated tissue probability maps, a sequence of local classifiers can be built for accurate tissue segmentation.

Original languageEnglish
Title of host publicationMedical Computer Vision and Bayesian and Graphical Models for Biomedical Imaging - MICCAI 2016 International Workshops, MCV and BAMBI, Revised Selected Papers
PublisherSpringer Verlag
Pages26-34
Number of pages9
Volume10081 LNCS
ISBN (Print)9783319611877
DOIs
Publication statusPublished - 2017
Externally publishedYes
EventInternational Workshop on Medical Computer Vision, MCV 2016, and of the International Workshop on Bayesian and grAphical Models for Biomedical Imaging, BAMBI 2016, held in conjunction with the 19th International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2016 - Athens, Greece
Duration: 2016 Oct 212016 Oct 21

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume10081 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Other

OtherInternational Workshop on Medical Computer Vision, MCV 2016, and of the International Workshop on Bayesian and grAphical Models for Biomedical Imaging, BAMBI 2016, held in conjunction with the 19th International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2016
CountryGreece
CityAthens
Period16/10/2116/10/21

Fingerprint

Brain
Classifiers
Segmentation
Classifier
Tissue
Voxel
Decision trees
Decision tree
Random Forest
Atlas
Training Samples
Magnetic resonance imaging
Refining
Training

ASJC Scopus subject areas

  • Theoretical Computer Science
  • Computer Science(all)

Cite this

Wang, L., Gao, Y., Li, G., Shi, F., Lin, W., & Shen, D. (2017). LATEST: Local AdapTivE and Sequential Training for tissue segmentation of isointense infant brain MR images. In Medical Computer Vision and Bayesian and Graphical Models for Biomedical Imaging - MICCAI 2016 International Workshops, MCV and BAMBI, Revised Selected Papers (Vol. 10081 LNCS, pp. 26-34). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 10081 LNCS). Springer Verlag. https://doi.org/10.1007/978-3-319-61188-4_3

LATEST : Local AdapTivE and Sequential Training for tissue segmentation of isointense infant brain MR images. / Wang, Li; Gao, Yaozong; Li, Gang; Shi, Feng; Lin, Weili; Shen, Dinggang.

Medical Computer Vision and Bayesian and Graphical Models for Biomedical Imaging - MICCAI 2016 International Workshops, MCV and BAMBI, Revised Selected Papers. Vol. 10081 LNCS Springer Verlag, 2017. p. 26-34 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 10081 LNCS).

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

Wang, L, Gao, Y, Li, G, Shi, F, Lin, W & Shen, D 2017, LATEST: Local AdapTivE and Sequential Training for tissue segmentation of isointense infant brain MR images. in Medical Computer Vision and Bayesian and Graphical Models for Biomedical Imaging - MICCAI 2016 International Workshops, MCV and BAMBI, Revised Selected Papers. vol. 10081 LNCS, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 10081 LNCS, Springer Verlag, pp. 26-34, International Workshop on Medical Computer Vision, MCV 2016, and of the International Workshop on Bayesian and grAphical Models for Biomedical Imaging, BAMBI 2016, held in conjunction with the 19th International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2016, Athens, Greece, 16/10/21. https://doi.org/10.1007/978-3-319-61188-4_3
Wang L, Gao Y, Li G, Shi F, Lin W, Shen D. LATEST: Local AdapTivE and Sequential Training for tissue segmentation of isointense infant brain MR images. In Medical Computer Vision and Bayesian and Graphical Models for Biomedical Imaging - MICCAI 2016 International Workshops, MCV and BAMBI, Revised Selected Papers. Vol. 10081 LNCS. Springer Verlag. 2017. p. 26-34. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). https://doi.org/10.1007/978-3-319-61188-4_3
Wang, Li ; Gao, Yaozong ; Li, Gang ; Shi, Feng ; Lin, Weili ; Shen, Dinggang. / LATEST : Local AdapTivE and Sequential Training for tissue segmentation of isointense infant brain MR images. Medical Computer Vision and Bayesian and Graphical Models for Biomedical Imaging - MICCAI 2016 International Workshops, MCV and BAMBI, Revised Selected Papers. Vol. 10081 LNCS Springer Verlag, 2017. pp. 26-34 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
@inproceedings{7396249bb38343079e72c5f1d7175d72,
title = "LATEST: Local AdapTivE and Sequential Training for tissue segmentation of isointense infant brain MR images",
abstract = "Accurate segmentation of isointense infant (~6 months of age) brain MRIs is of great importance, however, a very challenging task, due to extremely low tissue contrast caused by ongoing myelination processes. In this work, we propose a novel learning method based on Local AdapTivE and Sequential Training (LATEST) for segmentation. Specifically, random forest technique is employed to train a local classifier (a single decision tree) for each voxel in the common space based on the neighboring training samples from atlases. Then, for each given voxel, all trained nearby individual classifiers (decision trees) are grouped together to form a forest. Moreover, the estimated probabilities are further used as additional source images to train the next set of local classifiers for refining tissue classification. By iteratively training the subsequent classifiers based on the updated tissue probability maps, a sequence of local classifiers can be built for accurate tissue segmentation.",
author = "Li Wang and Yaozong Gao and Gang Li and Feng Shi and Weili Lin and Dinggang Shen",
year = "2017",
doi = "10.1007/978-3-319-61188-4_3",
language = "English",
isbn = "9783319611877",
volume = "10081 LNCS",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
publisher = "Springer Verlag",
pages = "26--34",
booktitle = "Medical Computer Vision and Bayesian and Graphical Models for Biomedical Imaging - MICCAI 2016 International Workshops, MCV and BAMBI, Revised Selected Papers",

}

TY - GEN

T1 - LATEST

T2 - Local AdapTivE and Sequential Training for tissue segmentation of isointense infant brain MR images

AU - Wang, Li

AU - Gao, Yaozong

AU - Li, Gang

AU - Shi, Feng

AU - Lin, Weili

AU - Shen, Dinggang

PY - 2017

Y1 - 2017

N2 - Accurate segmentation of isointense infant (~6 months of age) brain MRIs is of great importance, however, a very challenging task, due to extremely low tissue contrast caused by ongoing myelination processes. In this work, we propose a novel learning method based on Local AdapTivE and Sequential Training (LATEST) for segmentation. Specifically, random forest technique is employed to train a local classifier (a single decision tree) for each voxel in the common space based on the neighboring training samples from atlases. Then, for each given voxel, all trained nearby individual classifiers (decision trees) are grouped together to form a forest. Moreover, the estimated probabilities are further used as additional source images to train the next set of local classifiers for refining tissue classification. By iteratively training the subsequent classifiers based on the updated tissue probability maps, a sequence of local classifiers can be built for accurate tissue segmentation.

AB - Accurate segmentation of isointense infant (~6 months of age) brain MRIs is of great importance, however, a very challenging task, due to extremely low tissue contrast caused by ongoing myelination processes. In this work, we propose a novel learning method based on Local AdapTivE and Sequential Training (LATEST) for segmentation. Specifically, random forest technique is employed to train a local classifier (a single decision tree) for each voxel in the common space based on the neighboring training samples from atlases. Then, for each given voxel, all trained nearby individual classifiers (decision trees) are grouped together to form a forest. Moreover, the estimated probabilities are further used as additional source images to train the next set of local classifiers for refining tissue classification. By iteratively training the subsequent classifiers based on the updated tissue probability maps, a sequence of local classifiers can be built for accurate tissue segmentation.

UR - http://www.scopus.com/inward/record.url?scp=85025134939&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85025134939&partnerID=8YFLogxK

U2 - 10.1007/978-3-319-61188-4_3

DO - 10.1007/978-3-319-61188-4_3

M3 - Conference contribution

AN - SCOPUS:85025134939

SN - 9783319611877

VL - 10081 LNCS

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

SP - 26

EP - 34

BT - Medical Computer Vision and Bayesian and Graphical Models for Biomedical Imaging - MICCAI 2016 International Workshops, MCV and BAMBI, Revised Selected Papers

PB - Springer Verlag

ER -