A hybrid of deep network and hidden markov model for MCI identification with resting-state fMRI

Research output: Chapter in Book/Report/Conference proceedingChapter

9 Citations (Scopus)

Abstract

In this paper, we propose a novel method for modelling functional dynamics in resting-state fMRI (rs-fMRI) for Mild Cognitive Impairment (MCI) identification. Specifically, we devise a hybrid architecture by combining Deep Auto-Encoder (DAE) and Hidden Markov Model (HMM). The roles of DAE and HMM are, respectively, to discover hierarchical non-linear relations among features, by which we transform the original features into a lower dimension space, and to model dynamic characteristics inherent in rs-fMRI, i.e., internal state changes. By building a generative model with HMMs for each class individually, we estimate the data likelihood of a test subject as MCI or normal healthy control, based on which we identify the clinical label. In our experiments, we achieved the maximal accuracy of 81.08% with the proposed method, outperforming state-of-the-art methods in the literature.

Original languageEnglish
Title of host publicationLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
PublisherSpringer Verlag
Pages573-580
Number of pages8
Volume9349
DOIs
Publication statusPublished - 2015

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume9349
ISSN (Print)03029743
ISSN (Electronic)16113349

Fingerprint

Functional Magnetic Resonance Imaging
Hidden Markov models
Markov Model
Encoder
Labels
Dynamic models
Generative Models
Dynamic Characteristics
Likelihood
Transform
Internal
Experiments
Modeling
Estimate
Experiment
Magnetic Resonance Imaging
Model

ASJC Scopus subject areas

  • Computer Science(all)
  • Theoretical Computer Science

Cite this

Suk, H-I., Lee, S. W., & Shen, D. (2015). A hybrid of deep network and hidden markov model for MCI identification with resting-state fMRI. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 9349, pp. 573-580). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 9349). Springer Verlag. https://doi.org/10.1007/978-3-319-24553-9_70

A hybrid of deep network and hidden markov model for MCI identification with resting-state fMRI. / Suk, Heung-Il; Lee, Seong Whan; Shen, Dinggang.

Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol. 9349 Springer Verlag, 2015. p. 573-580 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 9349).

Research output: Chapter in Book/Report/Conference proceedingChapter

Suk, H-I, Lee, SW & Shen, D 2015, A hybrid of deep network and hidden markov model for MCI identification with resting-state fMRI. in Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). vol. 9349, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 9349, Springer Verlag, pp. 573-580. https://doi.org/10.1007/978-3-319-24553-9_70
Suk H-I, Lee SW, Shen D. A hybrid of deep network and hidden markov model for MCI identification with resting-state fMRI. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol. 9349. Springer Verlag. 2015. p. 573-580. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). https://doi.org/10.1007/978-3-319-24553-9_70
Suk, Heung-Il ; Lee, Seong Whan ; Shen, Dinggang. / A hybrid of deep network and hidden markov model for MCI identification with resting-state fMRI. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol. 9349 Springer Verlag, 2015. pp. 573-580 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
@inbook{69e55fc65bb54bfaabe83675ead4418a,
title = "A hybrid of deep network and hidden markov model for MCI identification with resting-state fMRI",
abstract = "In this paper, we propose a novel method for modelling functional dynamics in resting-state fMRI (rs-fMRI) for Mild Cognitive Impairment (MCI) identification. Specifically, we devise a hybrid architecture by combining Deep Auto-Encoder (DAE) and Hidden Markov Model (HMM). The roles of DAE and HMM are, respectively, to discover hierarchical non-linear relations among features, by which we transform the original features into a lower dimension space, and to model dynamic characteristics inherent in rs-fMRI, i.e., internal state changes. By building a generative model with HMMs for each class individually, we estimate the data likelihood of a test subject as MCI or normal healthy control, based on which we identify the clinical label. In our experiments, we achieved the maximal accuracy of 81.08{\%} with the proposed method, outperforming state-of-the-art methods in the literature.",
author = "Heung-Il Suk and Lee, {Seong Whan} and Dinggang Shen",
year = "2015",
doi = "10.1007/978-3-319-24553-9_70",
language = "English",
volume = "9349",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
publisher = "Springer Verlag",
pages = "573--580",
booktitle = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

}

TY - CHAP

T1 - A hybrid of deep network and hidden markov model for MCI identification with resting-state fMRI

AU - Suk, Heung-Il

AU - Lee, Seong Whan

AU - Shen, Dinggang

PY - 2015

Y1 - 2015

N2 - In this paper, we propose a novel method for modelling functional dynamics in resting-state fMRI (rs-fMRI) for Mild Cognitive Impairment (MCI) identification. Specifically, we devise a hybrid architecture by combining Deep Auto-Encoder (DAE) and Hidden Markov Model (HMM). The roles of DAE and HMM are, respectively, to discover hierarchical non-linear relations among features, by which we transform the original features into a lower dimension space, and to model dynamic characteristics inherent in rs-fMRI, i.e., internal state changes. By building a generative model with HMMs for each class individually, we estimate the data likelihood of a test subject as MCI or normal healthy control, based on which we identify the clinical label. In our experiments, we achieved the maximal accuracy of 81.08% with the proposed method, outperforming state-of-the-art methods in the literature.

AB - In this paper, we propose a novel method for modelling functional dynamics in resting-state fMRI (rs-fMRI) for Mild Cognitive Impairment (MCI) identification. Specifically, we devise a hybrid architecture by combining Deep Auto-Encoder (DAE) and Hidden Markov Model (HMM). The roles of DAE and HMM are, respectively, to discover hierarchical non-linear relations among features, by which we transform the original features into a lower dimension space, and to model dynamic characteristics inherent in rs-fMRI, i.e., internal state changes. By building a generative model with HMMs for each class individually, we estimate the data likelihood of a test subject as MCI or normal healthy control, based on which we identify the clinical label. In our experiments, we achieved the maximal accuracy of 81.08% with the proposed method, outperforming state-of-the-art methods in the literature.

UR - http://www.scopus.com/inward/record.url?scp=84947466198&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84947466198&partnerID=8YFLogxK

U2 - 10.1007/978-3-319-24553-9_70

DO - 10.1007/978-3-319-24553-9_70

M3 - Chapter

VL - 9349

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

SP - 573

EP - 580

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

PB - Springer Verlag

ER -