3D convolutional neural network for feature extraction and classification of fMRI volumes

Hanh Vu, Hyun Chul Kim, Jong-Hwan Lee

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

1 Citation (Scopus)

Abstract

Recently, deep learning (DL) techniques have been gaining interest in the neuroimaging community. In this study, we present 3D convolutional neural network (3D-CNN) as an end-To-end model to label a target task among four sensorimotor tasks for each functional magnetic resonance imaging (fMRI) volume. To the best of our knowledge, this is the first study that employs a single blood-oxygenation-level-dependent (BOLD) fMRI volume as the input of the 3D-CNN for task classification. We hypothesized that 3D-CNN has the capability to extract potentially shift-invariant features in local brain areas while preserving the overall spatial layout of the whole brain fMRI volume. We designed a 3D-CNN model by extending the LeNet-5 CNN for 2D image classification to 3D volume classification. The designed 3D-CNN model was thoroughly evaluated using BOLD fMRI volumes acquired from four sensorimotor tasks in terms of the classification performance and feature representations for each of the four sensorimotor tasks.

Original languageEnglish
Title of host publication2018 International Workshop on Pattern Recognition in Neuroimaging, PRNI 2018
PublisherInstitute of Electrical and Electronics Engineers Inc.
ISBN (Print)9781538668597
DOIs
Publication statusPublished - 2018 Jul 31
Event2018 International Workshop on Pattern Recognition in Neuroimaging, PRNI 2018 - Singapore, Singapore
Duration: 2018 Jun 122018 Jun 14

Other

Other2018 International Workshop on Pattern Recognition in Neuroimaging, PRNI 2018
CountrySingapore
CitySingapore
Period18/6/1218/6/14

Fingerprint

Feature extraction
Magnetic Resonance Imaging
Neural networks
Neural Networks (Computer)
Oxygenation
Brain
Blood
Neuroimaging
Image classification
Learning
Labels

Keywords

  • 3D-CNN
  • Convolutional neural network (CNN)
  • deep learning
  • fMRI
  • sensorimotor

ASJC Scopus subject areas

  • Computer Vision and Pattern Recognition
  • Radiology Nuclear Medicine and imaging
  • Behavioral Neuroscience
  • Cognitive Neuroscience
  • Neurology

Cite this

Vu, H., Kim, H. C., & Lee, J-H. (2018). 3D convolutional neural network for feature extraction and classification of fMRI volumes. In 2018 International Workshop on Pattern Recognition in Neuroimaging, PRNI 2018 [8423964] Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/PRNI.2018.8423964

3D convolutional neural network for feature extraction and classification of fMRI volumes. / Vu, Hanh; Kim, Hyun Chul; Lee, Jong-Hwan.

2018 International Workshop on Pattern Recognition in Neuroimaging, PRNI 2018. Institute of Electrical and Electronics Engineers Inc., 2018. 8423964.

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

Vu, H, Kim, HC & Lee, J-H 2018, 3D convolutional neural network for feature extraction and classification of fMRI volumes. in 2018 International Workshop on Pattern Recognition in Neuroimaging, PRNI 2018., 8423964, Institute of Electrical and Electronics Engineers Inc., 2018 International Workshop on Pattern Recognition in Neuroimaging, PRNI 2018, Singapore, Singapore, 18/6/12. https://doi.org/10.1109/PRNI.2018.8423964
Vu H, Kim HC, Lee J-H. 3D convolutional neural network for feature extraction and classification of fMRI volumes. In 2018 International Workshop on Pattern Recognition in Neuroimaging, PRNI 2018. Institute of Electrical and Electronics Engineers Inc. 2018. 8423964 https://doi.org/10.1109/PRNI.2018.8423964
Vu, Hanh ; Kim, Hyun Chul ; Lee, Jong-Hwan. / 3D convolutional neural network for feature extraction and classification of fMRI volumes. 2018 International Workshop on Pattern Recognition in Neuroimaging, PRNI 2018. Institute of Electrical and Electronics Engineers Inc., 2018.
@inproceedings{23c30faaff57440e822f0c7a516e29c3,
title = "3D convolutional neural network for feature extraction and classification of fMRI volumes",
abstract = "Recently, deep learning (DL) techniques have been gaining interest in the neuroimaging community. In this study, we present 3D convolutional neural network (3D-CNN) as an end-To-end model to label a target task among four sensorimotor tasks for each functional magnetic resonance imaging (fMRI) volume. To the best of our knowledge, this is the first study that employs a single blood-oxygenation-level-dependent (BOLD) fMRI volume as the input of the 3D-CNN for task classification. We hypothesized that 3D-CNN has the capability to extract potentially shift-invariant features in local brain areas while preserving the overall spatial layout of the whole brain fMRI volume. We designed a 3D-CNN model by extending the LeNet-5 CNN for 2D image classification to 3D volume classification. The designed 3D-CNN model was thoroughly evaluated using BOLD fMRI volumes acquired from four sensorimotor tasks in terms of the classification performance and feature representations for each of the four sensorimotor tasks.",
keywords = "3D-CNN, Convolutional neural network (CNN), deep learning, fMRI, sensorimotor",
author = "Hanh Vu and Kim, {Hyun Chul} and Jong-Hwan Lee",
year = "2018",
month = "7",
day = "31",
doi = "10.1109/PRNI.2018.8423964",
language = "English",
isbn = "9781538668597",
booktitle = "2018 International Workshop on Pattern Recognition in Neuroimaging, PRNI 2018",
publisher = "Institute of Electrical and Electronics Engineers Inc.",

}

TY - GEN

T1 - 3D convolutional neural network for feature extraction and classification of fMRI volumes

AU - Vu, Hanh

AU - Kim, Hyun Chul

AU - Lee, Jong-Hwan

PY - 2018/7/31

Y1 - 2018/7/31

N2 - Recently, deep learning (DL) techniques have been gaining interest in the neuroimaging community. In this study, we present 3D convolutional neural network (3D-CNN) as an end-To-end model to label a target task among four sensorimotor tasks for each functional magnetic resonance imaging (fMRI) volume. To the best of our knowledge, this is the first study that employs a single blood-oxygenation-level-dependent (BOLD) fMRI volume as the input of the 3D-CNN for task classification. We hypothesized that 3D-CNN has the capability to extract potentially shift-invariant features in local brain areas while preserving the overall spatial layout of the whole brain fMRI volume. We designed a 3D-CNN model by extending the LeNet-5 CNN for 2D image classification to 3D volume classification. The designed 3D-CNN model was thoroughly evaluated using BOLD fMRI volumes acquired from four sensorimotor tasks in terms of the classification performance and feature representations for each of the four sensorimotor tasks.

AB - Recently, deep learning (DL) techniques have been gaining interest in the neuroimaging community. In this study, we present 3D convolutional neural network (3D-CNN) as an end-To-end model to label a target task among four sensorimotor tasks for each functional magnetic resonance imaging (fMRI) volume. To the best of our knowledge, this is the first study that employs a single blood-oxygenation-level-dependent (BOLD) fMRI volume as the input of the 3D-CNN for task classification. We hypothesized that 3D-CNN has the capability to extract potentially shift-invariant features in local brain areas while preserving the overall spatial layout of the whole brain fMRI volume. We designed a 3D-CNN model by extending the LeNet-5 CNN for 2D image classification to 3D volume classification. The designed 3D-CNN model was thoroughly evaluated using BOLD fMRI volumes acquired from four sensorimotor tasks in terms of the classification performance and feature representations for each of the four sensorimotor tasks.

KW - 3D-CNN

KW - Convolutional neural network (CNN)

KW - deep learning

KW - fMRI

KW - sensorimotor

UR - http://www.scopus.com/inward/record.url?scp=85051527795&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85051527795&partnerID=8YFLogxK

U2 - 10.1109/PRNI.2018.8423964

DO - 10.1109/PRNI.2018.8423964

M3 - Conference contribution

AN - SCOPUS:85051527795

SN - 9781538668597

BT - 2018 International Workshop on Pattern Recognition in Neuroimaging, PRNI 2018

PB - Institute of Electrical and Electronics Engineers Inc.

ER -