Deep ensemble sparse regression network for alzheimer’s disease diagnosis

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

7 Citations (Scopus)

Abstract

For neuroimaging-based brain disease diagnosis, sparse regression models have proved their effectiveness in handling highdimensional data but with a small number of samples. In this paper, we propose a novel framework that utilizes sparse regression models as target-level representation learner and builds a deep convolutional neural network for clinical decision making. Specifically, we first train multiple sparse regression models, each of which has different values of a regularization control parameter, and use the outputs of the trained regression models as target-level representations. Note that sparse regression models trained with different values of a regularization control parameter potentially select different sets of features from the original ones, thereby they have different powers to predict the response values, i.e., a clinical label and clinical scores in our work. We then construct a deep convolutional neural network by taking the target-level representations as input. Our deep network learns to optimally fuse the predicted response variables, i.e., target-level representations, from the same sparse response model(s) and also those from the neighboring sparse response models. To our best knowledge, this is the first work that systematically integrates sparse regression models with deep neural network. In our experiments with ADNI cohort, we validated the effectiveness of the proposed method by achieving the highest classification accuracies in three different tasks of Alzheimer’s disease and mild cognitive impairment identification.

Original languageEnglish
Title of host publicationMachine Learning in Medical Imaging - 7th International Workshop, MLMI 2016 held in conjunction with MICCAI 2016, Proceedings
PublisherSpringer Verlag
Pages113-121
Number of pages9
Volume10019 LNCS
ISBN (Print)9783319471563
DOIs
Publication statusPublished - 2016
Event7th International Workshop on Machine Learning in Medical Imaging, MLMI 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)
Volume10019 LNCS
ISSN (Print)03029743
ISSN (Electronic)16113349

Other

Other7th International Workshop on Machine Learning in Medical Imaging, MLMI 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

Alzheimer's Disease
Ensemble
Regression
Regression Model
Target
Regularization Parameter
Neural Networks
Control Parameter
Neuroimaging
Data Handling
Neural networks
Data handling
Electric fuses
Labels
Decision Making
Integrate
Brain
Decision making
Predict
Output

ASJC Scopus subject areas

  • Theoretical Computer Science
  • Computer Science(all)

Cite this

Suk, H-I., & Shen, D. (2016). Deep ensemble sparse regression network for alzheimer’s disease diagnosis. In Machine Learning in Medical Imaging - 7th International Workshop, MLMI 2016 held in conjunction with MICCAI 2016, Proceedings (Vol. 10019 LNCS, pp. 113-121). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 10019 LNCS). Springer Verlag. https://doi.org/10.1007/978-3-319-47157-0_14

Deep ensemble sparse regression network for alzheimer’s disease diagnosis. / Suk, Heung-Il; Shen, Dinggang.

Machine Learning in Medical Imaging - 7th International Workshop, MLMI 2016 held in conjunction with MICCAI 2016, Proceedings. Vol. 10019 LNCS Springer Verlag, 2016. p. 113-121 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 10019 LNCS).

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

Suk, H-I & Shen, D 2016, Deep ensemble sparse regression network for alzheimer’s disease diagnosis. in Machine Learning in Medical Imaging - 7th International Workshop, MLMI 2016 held in conjunction with MICCAI 2016, Proceedings. vol. 10019 LNCS, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 10019 LNCS, Springer Verlag, pp. 113-121, 7th International Workshop on Machine Learning in Medical Imaging, MLMI 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/10.1007/978-3-319-47157-0_14
Suk H-I, Shen D. Deep ensemble sparse regression network for alzheimer’s disease diagnosis. In Machine Learning in Medical Imaging - 7th International Workshop, MLMI 2016 held in conjunction with MICCAI 2016, Proceedings. Vol. 10019 LNCS. Springer Verlag. 2016. p. 113-121. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). https://doi.org/10.1007/978-3-319-47157-0_14
Suk, Heung-Il ; Shen, Dinggang. / Deep ensemble sparse regression network for alzheimer’s disease diagnosis. Machine Learning in Medical Imaging - 7th International Workshop, MLMI 2016 held in conjunction with MICCAI 2016, Proceedings. Vol. 10019 LNCS Springer Verlag, 2016. pp. 113-121 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
@inproceedings{30fd71da7eb2434f93f7d1f55bf2a91c,
title = "Deep ensemble sparse regression network for alzheimer’s disease diagnosis",
abstract = "For neuroimaging-based brain disease diagnosis, sparse regression models have proved their effectiveness in handling highdimensional data but with a small number of samples. In this paper, we propose a novel framework that utilizes sparse regression models as target-level representation learner and builds a deep convolutional neural network for clinical decision making. Specifically, we first train multiple sparse regression models, each of which has different values of a regularization control parameter, and use the outputs of the trained regression models as target-level representations. Note that sparse regression models trained with different values of a regularization control parameter potentially select different sets of features from the original ones, thereby they have different powers to predict the response values, i.e., a clinical label and clinical scores in our work. We then construct a deep convolutional neural network by taking the target-level representations as input. Our deep network learns to optimally fuse the predicted response variables, i.e., target-level representations, from the same sparse response model(s) and also those from the neighboring sparse response models. To our best knowledge, this is the first work that systematically integrates sparse regression models with deep neural network. In our experiments with ADNI cohort, we validated the effectiveness of the proposed method by achieving the highest classification accuracies in three different tasks of Alzheimer’s disease and mild cognitive impairment identification.",
author = "Heung-Il Suk and Dinggang Shen",
year = "2016",
doi = "10.1007/978-3-319-47157-0_14",
language = "English",
isbn = "9783319471563",
volume = "10019 LNCS",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
publisher = "Springer Verlag",
pages = "113--121",
booktitle = "Machine Learning in Medical Imaging - 7th International Workshop, MLMI 2016 held in conjunction with MICCAI 2016, Proceedings",

}

TY - GEN

T1 - Deep ensemble sparse regression network for alzheimer’s disease diagnosis

AU - Suk, Heung-Il

AU - Shen, Dinggang

PY - 2016

Y1 - 2016

N2 - For neuroimaging-based brain disease diagnosis, sparse regression models have proved their effectiveness in handling highdimensional data but with a small number of samples. In this paper, we propose a novel framework that utilizes sparse regression models as target-level representation learner and builds a deep convolutional neural network for clinical decision making. Specifically, we first train multiple sparse regression models, each of which has different values of a regularization control parameter, and use the outputs of the trained regression models as target-level representations. Note that sparse regression models trained with different values of a regularization control parameter potentially select different sets of features from the original ones, thereby they have different powers to predict the response values, i.e., a clinical label and clinical scores in our work. We then construct a deep convolutional neural network by taking the target-level representations as input. Our deep network learns to optimally fuse the predicted response variables, i.e., target-level representations, from the same sparse response model(s) and also those from the neighboring sparse response models. To our best knowledge, this is the first work that systematically integrates sparse regression models with deep neural network. In our experiments with ADNI cohort, we validated the effectiveness of the proposed method by achieving the highest classification accuracies in three different tasks of Alzheimer’s disease and mild cognitive impairment identification.

AB - For neuroimaging-based brain disease diagnosis, sparse regression models have proved their effectiveness in handling highdimensional data but with a small number of samples. In this paper, we propose a novel framework that utilizes sparse regression models as target-level representation learner and builds a deep convolutional neural network for clinical decision making. Specifically, we first train multiple sparse regression models, each of which has different values of a regularization control parameter, and use the outputs of the trained regression models as target-level representations. Note that sparse regression models trained with different values of a regularization control parameter potentially select different sets of features from the original ones, thereby they have different powers to predict the response values, i.e., a clinical label and clinical scores in our work. We then construct a deep convolutional neural network by taking the target-level representations as input. Our deep network learns to optimally fuse the predicted response variables, i.e., target-level representations, from the same sparse response model(s) and also those from the neighboring sparse response models. To our best knowledge, this is the first work that systematically integrates sparse regression models with deep neural network. In our experiments with ADNI cohort, we validated the effectiveness of the proposed method by achieving the highest classification accuracies in three different tasks of Alzheimer’s disease and mild cognitive impairment identification.

UR - http://www.scopus.com/inward/record.url?scp=84992513254&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84992513254&partnerID=8YFLogxK

U2 - 10.1007/978-3-319-47157-0_14

DO - 10.1007/978-3-319-47157-0_14

M3 - Conference contribution

AN - SCOPUS:84992513254

SN - 9783319471563

VL - 10019 LNCS

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

SP - 113

EP - 121

BT - Machine Learning in Medical Imaging - 7th International Workshop, MLMI 2016 held in conjunction with MICCAI 2016, Proceedings

PB - Springer Verlag

ER -