Multi-source information gain for random forest: An application to CT image prediction from MRI data

Alzheimer’s Disease Neuroimaging Initiative (ADNI)

doi:10.1007/978-3-319-24888-2_39

Multi-source information gain for random forest: An application to CT image prediction from MRI data

Alzheimer’s Disease Neuroimaging Initiative (ADNI)

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

4 Citations (Scopus)

Abstract

Random forest has been widely recognized as one of the most powerful learning-based predictors in literature, with a broad range of applications in medical imaging. Notable efforts have been focused on enhancing the algorithm in multiple facets. In this paper, we present an original concept of multi-source information gain that escapes from the conventional notion inherent to random forest. We propose the idea of characterizing information gain in the training process by utilizing multiple beneficial sources of information, instead of the sole governing of prediction targets as conventionally known. We suggest the use of location and input image patches as the secondary sources of information for guiding the splitting process in random forest, and experiment on the challenging task of predicting CT images from MRI data. The experimentation is thoroughly analyzed in two datasets, i.e., human brain and prostate, with its performance further validated with the integration of auto-context model. Results prove that the multi-source information gain concept effectively helps better guide the training process with consistent improvement in prediction accuracy.

Original language	English
Title of host publication	Machine Learning in Medical Imaging - 6th International Workshop, MLMI 2015 Held in Conjunction with MICCAI 2015, Proceedings
Editors	Luping Zhou, Yinghuan Shi, Li Wang, Qian Wang
Publisher	Springer Verlag
Pages	321-329
Number of pages	9
ISBN (Print)	9783319248875
DOIs	https://doi.org/10.1007/978-3-319-24888-2_39
Publication status	Published - 2015
Event	6th International Workshop on Machine Learning in Medical Imaging, MLMI 2015 and Held in Conjunction with 18th International Conference on Medical Image Computing and Computer Assisted Intervention, MICCAI 2015 - Munich, Germany Duration: 2015 Oct 5 → 2015 Oct 5

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	9352
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Other

Other	6th International Workshop on Machine Learning in Medical Imaging, MLMI 2015 and Held in Conjunction with 18th International Conference on Medical Image Computing and Computer Assisted Intervention, MICCAI 2015
Country/Territory	Germany
City	Munich
Period	15/10/5 → 15/10/5

Bibliographical note

Publisher Copyright:
© Springer International Publishing Switzerland 2015.

ASJC Scopus subject areas

Theoretical Computer Science
General Computer Science

Access to Document

10.1007/978-3-319-24888-2_39

Cite this

Alzheimer’s Disease Neuroimaging Initiative (ADNI) (2015). Multi-source information gain for random forest: An application to CT image prediction from MRI data. In L. Zhou, Y. Shi, L. Wang, & Q. Wang (Eds.), Machine Learning in Medical Imaging - 6th International Workshop, MLMI 2015 Held in Conjunction with MICCAI 2015, Proceedings (pp. 321-329). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 9352). Springer Verlag. https://doi.org/10.1007/978-3-319-24888-2_39

Multi-source information gain for random forest: An application to CT image prediction from MRI data. / Alzheimer’s Disease Neuroimaging Initiative (ADNI).
Machine Learning in Medical Imaging - 6th International Workshop, MLMI 2015 Held in Conjunction with MICCAI 2015, Proceedings. ed. / Luping Zhou; Yinghuan Shi; Li Wang; Qian Wang. Springer Verlag, 2015. p. 321-329 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 9352).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Alzheimer’s Disease Neuroimaging Initiative (ADNI) 2015, Multi-source information gain for random forest: An application to CT image prediction from MRI data. in L Zhou, Y Shi, L Wang & Q Wang (eds), Machine Learning in Medical Imaging - 6th International Workshop, MLMI 2015 Held in Conjunction with MICCAI 2015, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 9352, Springer Verlag, pp. 321-329, 6th International Workshop on Machine Learning in Medical Imaging, MLMI 2015 and Held in Conjunction with 18th International Conference on Medical Image Computing and Computer Assisted Intervention, MICCAI 2015, Munich, Germany, 15/10/5. https://doi.org/10.1007/978-3-319-24888-2_39

Alzheimer’s Disease Neuroimaging Initiative (ADNI). Multi-source information gain for random forest: An application to CT image prediction from MRI data. In Zhou L, Shi Y, Wang L, Wang Q, editors, Machine Learning in Medical Imaging - 6th International Workshop, MLMI 2015 Held in Conjunction with MICCAI 2015, Proceedings. Springer Verlag. 2015. p. 321-329. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-319-24888-2_39

Alzheimer’s Disease Neuroimaging Initiative (ADNI). / Multi-source information gain for random forest : An application to CT image prediction from MRI data. Machine Learning in Medical Imaging - 6th International Workshop, MLMI 2015 Held in Conjunction with MICCAI 2015, Proceedings. editor / Luping Zhou ; Yinghuan Shi ; Li Wang ; Qian Wang. Springer Verlag, 2015. pp. 321-329 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{f735227753624ad8bceb561eb28bc6fa,

title = "Multi-source information gain for random forest: An application to CT image prediction from MRI data",

abstract = "Random forest has been widely recognized as one of the most powerful learning-based predictors in literature, with a broad range of applications in medical imaging. Notable efforts have been focused on enhancing the algorithm in multiple facets. In this paper, we present an original concept of multi-source information gain that escapes from the conventional notion inherent to random forest. We propose the idea of characterizing information gain in the training process by utilizing multiple beneficial sources of information, instead of the sole governing of prediction targets as conventionally known. We suggest the use of location and input image patches as the secondary sources of information for guiding the splitting process in random forest, and experiment on the challenging task of predicting CT images from MRI data. The experimentation is thoroughly analyzed in two datasets, i.e., human brain and prostate, with its performance further validated with the integration of auto-context model. Results prove that the multi-source information gain concept effectively helps better guide the training process with consistent improvement in prediction accuracy.",

author = "{Alzheimer{\textquoteright}s Disease Neuroimaging Initiative (ADNI)} and Tri Huynh and Yaozong Gao and Jiayin Kang and Li Wang and Pei Zhang and Dinggang Shen",

note = "Publisher Copyright: {\textcopyright} Springer International Publishing Switzerland 2015.; 6th International Workshop on Machine Learning in Medical Imaging, MLMI 2015 and Held in Conjunction with 18th International Conference on Medical Image Computing and Computer Assisted Intervention, MICCAI 2015 ; Conference date: 05-10-2015 Through 05-10-2015",

year = "2015",

doi = "10.1007/978-3-319-24888-2_39",

language = "English",

isbn = "9783319248875",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

publisher = "Springer Verlag",

pages = "321--329",

editor = "Luping Zhou and Yinghuan Shi and Li Wang and Qian Wang",

booktitle = "Machine Learning in Medical Imaging - 6th International Workshop, MLMI 2015 Held in Conjunction with MICCAI 2015, Proceedings",

}

TY - GEN

T1 - Multi-source information gain for random forest

T2 - 6th International Workshop on Machine Learning in Medical Imaging, MLMI 2015 and Held in Conjunction with 18th International Conference on Medical Image Computing and Computer Assisted Intervention, MICCAI 2015

AU - Alzheimer’s Disease Neuroimaging Initiative (ADNI)

AU - Huynh, Tri

AU - Gao, Yaozong

AU - Kang, Jiayin

AU - Wang, Li

AU - Zhang, Pei

AU - Shen, Dinggang

PY - 2015

Y1 - 2015

N2 - Random forest has been widely recognized as one of the most powerful learning-based predictors in literature, with a broad range of applications in medical imaging. Notable efforts have been focused on enhancing the algorithm in multiple facets. In this paper, we present an original concept of multi-source information gain that escapes from the conventional notion inherent to random forest. We propose the idea of characterizing information gain in the training process by utilizing multiple beneficial sources of information, instead of the sole governing of prediction targets as conventionally known. We suggest the use of location and input image patches as the secondary sources of information for guiding the splitting process in random forest, and experiment on the challenging task of predicting CT images from MRI data. The experimentation is thoroughly analyzed in two datasets, i.e., human brain and prostate, with its performance further validated with the integration of auto-context model. Results prove that the multi-source information gain concept effectively helps better guide the training process with consistent improvement in prediction accuracy.

AB - Random forest has been widely recognized as one of the most powerful learning-based predictors in literature, with a broad range of applications in medical imaging. Notable efforts have been focused on enhancing the algorithm in multiple facets. In this paper, we present an original concept of multi-source information gain that escapes from the conventional notion inherent to random forest. We propose the idea of characterizing information gain in the training process by utilizing multiple beneficial sources of information, instead of the sole governing of prediction targets as conventionally known. We suggest the use of location and input image patches as the secondary sources of information for guiding the splitting process in random forest, and experiment on the challenging task of predicting CT images from MRI data. The experimentation is thoroughly analyzed in two datasets, i.e., human brain and prostate, with its performance further validated with the integration of auto-context model. Results prove that the multi-source information gain concept effectively helps better guide the training process with consistent improvement in prediction accuracy.

UR - http://www.scopus.com/inward/record.url?scp=84951985142&partnerID=8YFLogxK

U2 - 10.1007/978-3-319-24888-2_39

DO - 10.1007/978-3-319-24888-2_39

M3 - Conference contribution

AN - SCOPUS:84951985142

SN - 9783319248875

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

SP - 321

EP - 329

BT - Machine Learning in Medical Imaging - 6th International Workshop, MLMI 2015 Held in Conjunction with MICCAI 2015, Proceedings

A2 - Zhou, Luping

A2 - Shi, Yinghuan

A2 - Wang, Li

A2 - Wang, Qian

PB - Springer Verlag

Y2 - 5 October 2015 through 5 October 2015

ER -

Multi-source information gain for random forest: An application to CT image prediction from MRI data

Abstract

Publication series

Other

Bibliographical note

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this