Classification algorithms using multiple MRI features in mild traumatic brain injury

Yvonne W. Lui, Yuanyi Xue, Damon Kenul, Yulin Ge, Robert I. Grossman, Yao Wang

Research output: Contribution to journalArticle

Abstract

Objective: The purpose of this study was to develop an algorithm incorporating MRI metrics to classify patients with mild traumatic brain injury (mTBI) and controls. Methods: This was an institutional review board-approved, Health Insurance Portability and Accountability Act-compliant prospective study. We recruited patients with mTBI and healthy controls through the emergency department and general population. We acquired data on a 3.0T Siemens Trio magnet including conventional brain imaging, resting-state fMRI, diffusionweighted imaging, and magnetic field correlation (MFC), and performed multifeature analysis using the following MRI metrics: mean kurtosis (MK) of thalamus, MFC of thalamus and frontal white matter, thalamocortical resting-state networks, and = regional gray matter and white matter volumes including the anterior cingulum and left frontal and temporal poles. Feature selection was performed using minimal-redundancy maximal-relevance. We used classifiers including support vector machine, naive Bayesian, Bayesian network, radial basis network, and multilayer perceptron to test maximal accuracy. Results: We studied 24 patients with mTBI and 26 controls. Best single-feature classification uses thalamic MK yielding 74% accuracy. Multifeature analysis yields 80% accuracy using the full feature set, and up to 86% accuracy using minimal-redundancy maximal-relevance feature selection (MK thalamus, right anterior cingulate volume, thalamic thickness, thalamocortical resting-state network, thalamic microscopic MFC, and sex). Conclusion: Multifeature analysis using diffusion-weighted imaging, MFC, fMRI, and volumetrics may aid in the classification of patients with mTBI compared with controls based on optimal feature selection and classification methods. Classification of evidence: This study provides Class III evidence that classification algorithms using multiple MRI features accurately identifies patients with mTBI as defined by American Congress of Rehabilitation Medicine criteria compared with healthy controls.

Original languageEnglish (US)
Pages (from-to)1235-1240
Number of pages6
JournalNeurology
Volume83
Issue number14
DOIs
StatePublished - Sep 1 2014

Fingerprint

Brain Concussion
Magnetic Fields
Thalamus
Magnetic Resonance Imaging
Health Insurance Portability and Accountability Act
Neural Networks (Computer)
Magnets
Research Ethics Committees
Gyrus Cinguli
Neuroimaging
Hospital Emergency Service
Rehabilitation
Medicine
Prospective Studies
Population

ASJC Scopus subject areas

  • Clinical Neurology
  • Medicine(all)

Cite this

Classification algorithms using multiple MRI features in mild traumatic brain injury. / Lui, Yvonne W.; Xue, Yuanyi; Kenul, Damon; Ge, Yulin; Grossman, Robert I.; Wang, Yao.

In: Neurology, Vol. 83, No. 14, 01.09.2014, p. 1235-1240.

Research output: Contribution to journalArticle

Lui, Yvonne W. ; Xue, Yuanyi ; Kenul, Damon ; Ge, Yulin ; Grossman, Robert I. ; Wang, Yao. / Classification algorithms using multiple MRI features in mild traumatic brain injury. In: Neurology. 2014 ; Vol. 83, No. 14. pp. 1235-1240.
@article{448c104e36714ef583256877b4fb060f,
title = "Classification algorithms using multiple MRI features in mild traumatic brain injury",
abstract = "Objective: The purpose of this study was to develop an algorithm incorporating MRI metrics to classify patients with mild traumatic brain injury (mTBI) and controls. Methods: This was an institutional review board-approved, Health Insurance Portability and Accountability Act-compliant prospective study. We recruited patients with mTBI and healthy controls through the emergency department and general population. We acquired data on a 3.0T Siemens Trio magnet including conventional brain imaging, resting-state fMRI, diffusionweighted imaging, and magnetic field correlation (MFC), and performed multifeature analysis using the following MRI metrics: mean kurtosis (MK) of thalamus, MFC of thalamus and frontal white matter, thalamocortical resting-state networks, and = regional gray matter and white matter volumes including the anterior cingulum and left frontal and temporal poles. Feature selection was performed using minimal-redundancy maximal-relevance. We used classifiers including support vector machine, naive Bayesian, Bayesian network, radial basis network, and multilayer perceptron to test maximal accuracy. Results: We studied 24 patients with mTBI and 26 controls. Best single-feature classification uses thalamic MK yielding 74{\%} accuracy. Multifeature analysis yields 80{\%} accuracy using the full feature set, and up to 86{\%} accuracy using minimal-redundancy maximal-relevance feature selection (MK thalamus, right anterior cingulate volume, thalamic thickness, thalamocortical resting-state network, thalamic microscopic MFC, and sex). Conclusion: Multifeature analysis using diffusion-weighted imaging, MFC, fMRI, and volumetrics may aid in the classification of patients with mTBI compared with controls based on optimal feature selection and classification methods. Classification of evidence: This study provides Class III evidence that classification algorithms using multiple MRI features accurately identifies patients with mTBI as defined by American Congress of Rehabilitation Medicine criteria compared with healthy controls.",
author = "Lui, {Yvonne W.} and Yuanyi Xue and Damon Kenul and Yulin Ge and Grossman, {Robert I.} and Yao Wang",
year = "2014",
month = "9",
day = "1",
doi = "10.1212/WNL.0000000000000834",
language = "English (US)",
volume = "83",
pages = "1235--1240",
journal = "Neurology",
issn = "0028-3878",
publisher = "Lippincott Williams and Wilkins",
number = "14",

}

TY - JOUR

T1 - Classification algorithms using multiple MRI features in mild traumatic brain injury

AU - Lui, Yvonne W.

AU - Xue, Yuanyi

AU - Kenul, Damon

AU - Ge, Yulin

AU - Grossman, Robert I.

AU - Wang, Yao

PY - 2014/9/1

Y1 - 2014/9/1

N2 - Objective: The purpose of this study was to develop an algorithm incorporating MRI metrics to classify patients with mild traumatic brain injury (mTBI) and controls. Methods: This was an institutional review board-approved, Health Insurance Portability and Accountability Act-compliant prospective study. We recruited patients with mTBI and healthy controls through the emergency department and general population. We acquired data on a 3.0T Siemens Trio magnet including conventional brain imaging, resting-state fMRI, diffusionweighted imaging, and magnetic field correlation (MFC), and performed multifeature analysis using the following MRI metrics: mean kurtosis (MK) of thalamus, MFC of thalamus and frontal white matter, thalamocortical resting-state networks, and = regional gray matter and white matter volumes including the anterior cingulum and left frontal and temporal poles. Feature selection was performed using minimal-redundancy maximal-relevance. We used classifiers including support vector machine, naive Bayesian, Bayesian network, radial basis network, and multilayer perceptron to test maximal accuracy. Results: We studied 24 patients with mTBI and 26 controls. Best single-feature classification uses thalamic MK yielding 74% accuracy. Multifeature analysis yields 80% accuracy using the full feature set, and up to 86% accuracy using minimal-redundancy maximal-relevance feature selection (MK thalamus, right anterior cingulate volume, thalamic thickness, thalamocortical resting-state network, thalamic microscopic MFC, and sex). Conclusion: Multifeature analysis using diffusion-weighted imaging, MFC, fMRI, and volumetrics may aid in the classification of patients with mTBI compared with controls based on optimal feature selection and classification methods. Classification of evidence: This study provides Class III evidence that classification algorithms using multiple MRI features accurately identifies patients with mTBI as defined by American Congress of Rehabilitation Medicine criteria compared with healthy controls.

AB - Objective: The purpose of this study was to develop an algorithm incorporating MRI metrics to classify patients with mild traumatic brain injury (mTBI) and controls. Methods: This was an institutional review board-approved, Health Insurance Portability and Accountability Act-compliant prospective study. We recruited patients with mTBI and healthy controls through the emergency department and general population. We acquired data on a 3.0T Siemens Trio magnet including conventional brain imaging, resting-state fMRI, diffusionweighted imaging, and magnetic field correlation (MFC), and performed multifeature analysis using the following MRI metrics: mean kurtosis (MK) of thalamus, MFC of thalamus and frontal white matter, thalamocortical resting-state networks, and = regional gray matter and white matter volumes including the anterior cingulum and left frontal and temporal poles. Feature selection was performed using minimal-redundancy maximal-relevance. We used classifiers including support vector machine, naive Bayesian, Bayesian network, radial basis network, and multilayer perceptron to test maximal accuracy. Results: We studied 24 patients with mTBI and 26 controls. Best single-feature classification uses thalamic MK yielding 74% accuracy. Multifeature analysis yields 80% accuracy using the full feature set, and up to 86% accuracy using minimal-redundancy maximal-relevance feature selection (MK thalamus, right anterior cingulate volume, thalamic thickness, thalamocortical resting-state network, thalamic microscopic MFC, and sex). Conclusion: Multifeature analysis using diffusion-weighted imaging, MFC, fMRI, and volumetrics may aid in the classification of patients with mTBI compared with controls based on optimal feature selection and classification methods. Classification of evidence: This study provides Class III evidence that classification algorithms using multiple MRI features accurately identifies patients with mTBI as defined by American Congress of Rehabilitation Medicine criteria compared with healthy controls.

UR - http://www.scopus.com/inward/record.url?scp=84921770175&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84921770175&partnerID=8YFLogxK

U2 - 10.1212/WNL.0000000000000834

DO - 10.1212/WNL.0000000000000834

M3 - Article

C2 - 25171930

AN - SCOPUS:84921770175

VL - 83

SP - 1235

EP - 1240

JO - Neurology

JF - Neurology

SN - 0028-3878

IS - 14

ER -