Quantifying the impact of type 2 diabetes on brain perfusion using deep neural networks

Behrouz Saghafi, Prabhat Garg, Benjamin C. Wagner, S. Carrie Smith, Jianzhao Xu, Ananth J. Madhuranthakam, Youngkyoo Jung, Jasmin Divers, Barry I. Freedman, Joseph A. Maldjian, Albert Montillo

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

Abstract

The effect of Type 2 Diabetes (T2D) on brain health is poorly understood. This study aims to quantify the association between T2D and perfusion in the brain. T2D is a very common metabolic disorder that can cause long term damage to the renal and cardiovascular systems. Previous research has discovered the shape, volume and white matter microstructures in the brain to be significantly impacted by T2D. We propose a fully-connected deep neural network to classify the regional Cerebral Blood Flow into low or high levels, given 16 clinical measures as predictors. The clinical measures include diabetes, renal, cardiovascular and demographics measures. Our model enables us to discover any nonlinear association which might exist between the input features and target. Moreover, our end-to-end architecture automatically learns the most relevant features and combines them without the need for applying a feature selection method. We achieved promising classification performance. Furthermore, in comparison with six (6) classical machine learning algorithms and six (6) alternative deep neural networks similarly tuned for the task, our proposed model outperformed all of them.

Original languageEnglish (US)
Title of host publicationDeep Learning in Medical Image Analysis and Multimodal Learning for Clinical Decision Support - 3rd International Workshop, DLMIA 2017 and 7th International Workshop, ML-CDS 2017 Held in Conjunction with MICCAI 2017, Proceedings
EditorsTal Arbel, M. Jorge Cardoso
PublisherSpringer-Verlag
Pages151-159
Number of pages9
ISBN (Print)9783319675572
DOIs
StatePublished - Jan 1 2017
Externally publishedYes
Event3rd International Workshop on Deep Learning in Medical Image Analysis, DLMIA 2017 and 7th International Workshop on Multimodal Learning for Clinical Decision Support, ML-CDS 2017 held in Conjunction with 20th International Conference on Medical Image Computing and Computer Assisted Intervention, MICCAI 2017 - Quebec City, Canada
Duration: Sep 14 2017Sep 14 2017

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume10553 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference3rd International Workshop on Deep Learning in Medical Image Analysis, DLMIA 2017 and 7th International Workshop on Multimodal Learning for Clinical Decision Support, ML-CDS 2017 held in Conjunction with 20th International Conference on Medical Image Computing and Computer Assisted Intervention, MICCAI 2017
CountryCanada
CityQuebec City
Period9/14/179/14/17

Fingerprint

Diabetes
Medical problems
Brain
Neural Networks
Cardiovascular System
Cardiovascular system
Blood Flow
Feature Selection
Learning algorithms
Learning systems
Disorder
Feature extraction
Learning Algorithm
Microstructure
Predictors
Machine Learning
Health
Quantify
Blood
Damage

ASJC Scopus subject areas

  • Theoretical Computer Science
  • Computer Science(all)

Cite this

Saghafi, B., Garg, P., Wagner, B. C., Smith, S. C., Xu, J., Madhuranthakam, A. J., ... Montillo, A. (2017). Quantifying the impact of type 2 diabetes on brain perfusion using deep neural networks. In T. Arbel, & M. J. Cardoso (Eds.), Deep Learning in Medical Image Analysis and Multimodal Learning for Clinical Decision Support - 3rd International Workshop, DLMIA 2017 and 7th International Workshop, ML-CDS 2017 Held in Conjunction with MICCAI 2017, Proceedings (pp. 151-159). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 10553 LNCS). Springer-Verlag. https://doi.org/10.1007/978-3-319-67558-9_18

Quantifying the impact of type 2 diabetes on brain perfusion using deep neural networks. / Saghafi, Behrouz; Garg, Prabhat; Wagner, Benjamin C.; Smith, S. Carrie; Xu, Jianzhao; Madhuranthakam, Ananth J.; Jung, Youngkyoo; Divers, Jasmin; Freedman, Barry I.; Maldjian, Joseph A.; Montillo, Albert.

Deep Learning in Medical Image Analysis and Multimodal Learning for Clinical Decision Support - 3rd International Workshop, DLMIA 2017 and 7th International Workshop, ML-CDS 2017 Held in Conjunction with MICCAI 2017, Proceedings. ed. / Tal Arbel; M. Jorge Cardoso. Springer-Verlag, 2017. p. 151-159 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 10553 LNCS).

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

Saghafi, B, Garg, P, Wagner, BC, Smith, SC, Xu, J, Madhuranthakam, AJ, Jung, Y, Divers, J, Freedman, BI, Maldjian, JA & Montillo, A 2017, Quantifying the impact of type 2 diabetes on brain perfusion using deep neural networks. in T Arbel & MJ Cardoso (eds), Deep Learning in Medical Image Analysis and Multimodal Learning for Clinical Decision Support - 3rd International Workshop, DLMIA 2017 and 7th International Workshop, ML-CDS 2017 Held in Conjunction with MICCAI 2017, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 10553 LNCS, Springer-Verlag, pp. 151-159, 3rd International Workshop on Deep Learning in Medical Image Analysis, DLMIA 2017 and 7th International Workshop on Multimodal Learning for Clinical Decision Support, ML-CDS 2017 held in Conjunction with 20th International Conference on Medical Image Computing and Computer Assisted Intervention, MICCAI 2017, Quebec City, Canada, 9/14/17. https://doi.org/10.1007/978-3-319-67558-9_18
Saghafi B, Garg P, Wagner BC, Smith SC, Xu J, Madhuranthakam AJ et al. Quantifying the impact of type 2 diabetes on brain perfusion using deep neural networks. In Arbel T, Cardoso MJ, editors, Deep Learning in Medical Image Analysis and Multimodal Learning for Clinical Decision Support - 3rd International Workshop, DLMIA 2017 and 7th International Workshop, ML-CDS 2017 Held in Conjunction with MICCAI 2017, Proceedings. Springer-Verlag. 2017. p. 151-159. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). https://doi.org/10.1007/978-3-319-67558-9_18
Saghafi, Behrouz ; Garg, Prabhat ; Wagner, Benjamin C. ; Smith, S. Carrie ; Xu, Jianzhao ; Madhuranthakam, Ananth J. ; Jung, Youngkyoo ; Divers, Jasmin ; Freedman, Barry I. ; Maldjian, Joseph A. ; Montillo, Albert. / Quantifying the impact of type 2 diabetes on brain perfusion using deep neural networks. Deep Learning in Medical Image Analysis and Multimodal Learning for Clinical Decision Support - 3rd International Workshop, DLMIA 2017 and 7th International Workshop, ML-CDS 2017 Held in Conjunction with MICCAI 2017, Proceedings. editor / Tal Arbel ; M. Jorge Cardoso. Springer-Verlag, 2017. pp. 151-159 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
@inproceedings{a372c0cbe2d941508e889e6c35913c2c,
title = "Quantifying the impact of type 2 diabetes on brain perfusion using deep neural networks",
abstract = "The effect of Type 2 Diabetes (T2D) on brain health is poorly understood. This study aims to quantify the association between T2D and perfusion in the brain. T2D is a very common metabolic disorder that can cause long term damage to the renal and cardiovascular systems. Previous research has discovered the shape, volume and white matter microstructures in the brain to be significantly impacted by T2D. We propose a fully-connected deep neural network to classify the regional Cerebral Blood Flow into low or high levels, given 16 clinical measures as predictors. The clinical measures include diabetes, renal, cardiovascular and demographics measures. Our model enables us to discover any nonlinear association which might exist between the input features and target. Moreover, our end-to-end architecture automatically learns the most relevant features and combines them without the need for applying a feature selection method. We achieved promising classification performance. Furthermore, in comparison with six (6) classical machine learning algorithms and six (6) alternative deep neural networks similarly tuned for the task, our proposed model outperformed all of them.",
author = "Behrouz Saghafi and Prabhat Garg and Wagner, {Benjamin C.} and Smith, {S. Carrie} and Jianzhao Xu and Madhuranthakam, {Ananth J.} and Youngkyoo Jung and Jasmin Divers and Freedman, {Barry I.} and Maldjian, {Joseph A.} and Albert Montillo",
year = "2017",
month = "1",
day = "1",
doi = "10.1007/978-3-319-67558-9_18",
language = "English (US)",
isbn = "9783319675572",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
publisher = "Springer-Verlag",
pages = "151--159",
editor = "Tal Arbel and Cardoso, {M. Jorge}",
booktitle = "Deep Learning in Medical Image Analysis and Multimodal Learning for Clinical Decision Support - 3rd International Workshop, DLMIA 2017 and 7th International Workshop, ML-CDS 2017 Held in Conjunction with MICCAI 2017, Proceedings",

}

TY - GEN

T1 - Quantifying the impact of type 2 diabetes on brain perfusion using deep neural networks

AU - Saghafi, Behrouz

AU - Garg, Prabhat

AU - Wagner, Benjamin C.

AU - Smith, S. Carrie

AU - Xu, Jianzhao

AU - Madhuranthakam, Ananth J.

AU - Jung, Youngkyoo

AU - Divers, Jasmin

AU - Freedman, Barry I.

AU - Maldjian, Joseph A.

AU - Montillo, Albert

PY - 2017/1/1

Y1 - 2017/1/1

N2 - The effect of Type 2 Diabetes (T2D) on brain health is poorly understood. This study aims to quantify the association between T2D and perfusion in the brain. T2D is a very common metabolic disorder that can cause long term damage to the renal and cardiovascular systems. Previous research has discovered the shape, volume and white matter microstructures in the brain to be significantly impacted by T2D. We propose a fully-connected deep neural network to classify the regional Cerebral Blood Flow into low or high levels, given 16 clinical measures as predictors. The clinical measures include diabetes, renal, cardiovascular and demographics measures. Our model enables us to discover any nonlinear association which might exist between the input features and target. Moreover, our end-to-end architecture automatically learns the most relevant features and combines them without the need for applying a feature selection method. We achieved promising classification performance. Furthermore, in comparison with six (6) classical machine learning algorithms and six (6) alternative deep neural networks similarly tuned for the task, our proposed model outperformed all of them.

AB - The effect of Type 2 Diabetes (T2D) on brain health is poorly understood. This study aims to quantify the association between T2D and perfusion in the brain. T2D is a very common metabolic disorder that can cause long term damage to the renal and cardiovascular systems. Previous research has discovered the shape, volume and white matter microstructures in the brain to be significantly impacted by T2D. We propose a fully-connected deep neural network to classify the regional Cerebral Blood Flow into low or high levels, given 16 clinical measures as predictors. The clinical measures include diabetes, renal, cardiovascular and demographics measures. Our model enables us to discover any nonlinear association which might exist between the input features and target. Moreover, our end-to-end architecture automatically learns the most relevant features and combines them without the need for applying a feature selection method. We achieved promising classification performance. Furthermore, in comparison with six (6) classical machine learning algorithms and six (6) alternative deep neural networks similarly tuned for the task, our proposed model outperformed all of them.

UR - http://www.scopus.com/inward/record.url?scp=85029813727&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85029813727&partnerID=8YFLogxK

U2 - 10.1007/978-3-319-67558-9_18

DO - 10.1007/978-3-319-67558-9_18

M3 - Conference contribution

AN - SCOPUS:85029813727

SN - 9783319675572

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

SP - 151

EP - 159

BT - Deep Learning in Medical Image Analysis and Multimodal Learning for Clinical Decision Support - 3rd International Workshop, DLMIA 2017 and 7th International Workshop, ML-CDS 2017 Held in Conjunction with MICCAI 2017, Proceedings

A2 - Arbel, Tal

A2 - Cardoso, M. Jorge

PB - Springer-Verlag

ER -