Optimizing Machine Learning Methods to Improve Predictive Models of Alzheimer's Disease

Alzheimer's Disease Neuroimaging Initiative

Research output: Contribution to journalArticle

Abstract

Background: Predicting clinical course of cognitive decline can boost clinical trials' power and improve our clinical decision-making. Machine learning (ML) algorithms are specifically designed for the purpose of prediction; however. identifying optimal features or algorithms is still a challenge. Objective: To investigate the accuracy of different ML methods and different features to classify cognitively normal (CN) individuals from Alzheimer's disease (AD) and to predict longitudinal outcome in participants with mild cognitive impairment (MCI). Methods: A total of 1,329 participants from the Alzheimer's Disease Neuroimaging Initiative (ADNI) were included: 424 CN, 656 MCI, and 249 AD individuals. Four feature-sets at baseline (hippocampal volume and volume of 47 cortical and subcortical regions with and without demographics and APOE4) and six machine learning methods (decision trees, support vector machines, K-nearest neighbor, ensemble linear discriminant, boosted trees, and random forests) were used to classify participants with normal cognition from participants with AD. Subsequently the model with best classification performance was used for predicting clinical outcome of MCI participants. Results: Ensemble linear discriminant models using demographics and all volumetric magnetic resonance imaging measures as feature-set showed the best performance in classification of CN versus AD participants (accuracy = 92.8%, sensitivity = 95.8%, and specificity = 88.3%). Prediction accuracy of future conversion from MCI to AD for this ensemble linear discriminant at 6, 12, 24, 36, and 48 months was 63.8% (sensitivity = 74.4, specificity = 63.1), 68.9% (sensitivity = 75.9, specificity = 67.8), 74.9% (sensitivity = 71.5, specificity = 76.3), 75.3%, (sensitivity = 65.2, specificity = 79.7), and 77.0% (sensitivity = 59.6, specificity = 86.1), respectively. Conclusions: Machine learning models trained for classification of CN versus AD can improve our prediction ability of MCI conversion to AD.

Original languageEnglish (US)
Pages (from-to)1027-1036
Number of pages10
JournalJournal of Alzheimer's Disease
Volume71
Issue number3
DOIs
StatePublished - Jan 1 2019

Fingerprint

Alzheimer Disease
Sensitivity and Specificity
Demography
Decision Trees
Aptitude
Machine Learning
Neuroimaging
Cognition
Cognitive Dysfunction
Linear Models
Magnetic Resonance Imaging
Clinical Trials

Keywords

  • Alzheimer's disease
  • classification
  • early diagnosis
  • machine learning
  • magnetic resonance imaging
  • mild cognitive impairment
  • predictive analytics

ASJC Scopus subject areas

  • Neuroscience(all)
  • Clinical Psychology
  • Geriatrics and Gerontology
  • Psychiatry and Mental health

Cite this

Optimizing Machine Learning Methods to Improve Predictive Models of Alzheimer's Disease. / Alzheimer's Disease Neuroimaging Initiative.

In: Journal of Alzheimer's Disease, Vol. 71, No. 3, 01.01.2019, p. 1027-1036.

Research output: Contribution to journalArticle

Alzheimer's Disease Neuroimaging Initiative. / Optimizing Machine Learning Methods to Improve Predictive Models of Alzheimer's Disease. In: Journal of Alzheimer's Disease. 2019 ; Vol. 71, No. 3. pp. 1027-1036.
@article{adac24e22a834c378e8397035f61c211,
title = "Optimizing Machine Learning Methods to Improve Predictive Models of Alzheimer's Disease",
abstract = "Background: Predicting clinical course of cognitive decline can boost clinical trials' power and improve our clinical decision-making. Machine learning (ML) algorithms are specifically designed for the purpose of prediction; however. identifying optimal features or algorithms is still a challenge. Objective: To investigate the accuracy of different ML methods and different features to classify cognitively normal (CN) individuals from Alzheimer's disease (AD) and to predict longitudinal outcome in participants with mild cognitive impairment (MCI). Methods: A total of 1,329 participants from the Alzheimer's Disease Neuroimaging Initiative (ADNI) were included: 424 CN, 656 MCI, and 249 AD individuals. Four feature-sets at baseline (hippocampal volume and volume of 47 cortical and subcortical regions with and without demographics and APOE4) and six machine learning methods (decision trees, support vector machines, K-nearest neighbor, ensemble linear discriminant, boosted trees, and random forests) were used to classify participants with normal cognition from participants with AD. Subsequently the model with best classification performance was used for predicting clinical outcome of MCI participants. Results: Ensemble linear discriminant models using demographics and all volumetric magnetic resonance imaging measures as feature-set showed the best performance in classification of CN versus AD participants (accuracy = 92.8{\%}, sensitivity = 95.8{\%}, and specificity = 88.3{\%}). Prediction accuracy of future conversion from MCI to AD for this ensemble linear discriminant at 6, 12, 24, 36, and 48 months was 63.8{\%} (sensitivity = 74.4, specificity = 63.1), 68.9{\%} (sensitivity = 75.9, specificity = 67.8), 74.9{\%} (sensitivity = 71.5, specificity = 76.3), 75.3{\%}, (sensitivity = 65.2, specificity = 79.7), and 77.0{\%} (sensitivity = 59.6, specificity = 86.1), respectively. Conclusions: Machine learning models trained for classification of CN versus AD can improve our prediction ability of MCI conversion to AD.",
keywords = "Alzheimer's disease, classification, early diagnosis, machine learning, magnetic resonance imaging, mild cognitive impairment, predictive analytics",
author = "{Alzheimer's Disease Neuroimaging Initiative} and Ali Ezzati and Zammit, {Andrea R.} and Harvey, {Danielle J.} and Christian Habeck and Hall, {Charles B.} and Lipton, {Richard B.}",
year = "2019",
month = "1",
day = "1",
doi = "10.3233/JAD-190262",
language = "English (US)",
volume = "71",
pages = "1027--1036",
journal = "Journal of Alzheimer's Disease",
issn = "1387-2877",
publisher = "IOS Press",
number = "3",

}

TY - JOUR

T1 - Optimizing Machine Learning Methods to Improve Predictive Models of Alzheimer's Disease

AU - Alzheimer's Disease Neuroimaging Initiative

AU - Ezzati, Ali

AU - Zammit, Andrea R.

AU - Harvey, Danielle J.

AU - Habeck, Christian

AU - Hall, Charles B.

AU - Lipton, Richard B.

PY - 2019/1/1

Y1 - 2019/1/1

N2 - Background: Predicting clinical course of cognitive decline can boost clinical trials' power and improve our clinical decision-making. Machine learning (ML) algorithms are specifically designed for the purpose of prediction; however. identifying optimal features or algorithms is still a challenge. Objective: To investigate the accuracy of different ML methods and different features to classify cognitively normal (CN) individuals from Alzheimer's disease (AD) and to predict longitudinal outcome in participants with mild cognitive impairment (MCI). Methods: A total of 1,329 participants from the Alzheimer's Disease Neuroimaging Initiative (ADNI) were included: 424 CN, 656 MCI, and 249 AD individuals. Four feature-sets at baseline (hippocampal volume and volume of 47 cortical and subcortical regions with and without demographics and APOE4) and six machine learning methods (decision trees, support vector machines, K-nearest neighbor, ensemble linear discriminant, boosted trees, and random forests) were used to classify participants with normal cognition from participants with AD. Subsequently the model with best classification performance was used for predicting clinical outcome of MCI participants. Results: Ensemble linear discriminant models using demographics and all volumetric magnetic resonance imaging measures as feature-set showed the best performance in classification of CN versus AD participants (accuracy = 92.8%, sensitivity = 95.8%, and specificity = 88.3%). Prediction accuracy of future conversion from MCI to AD for this ensemble linear discriminant at 6, 12, 24, 36, and 48 months was 63.8% (sensitivity = 74.4, specificity = 63.1), 68.9% (sensitivity = 75.9, specificity = 67.8), 74.9% (sensitivity = 71.5, specificity = 76.3), 75.3%, (sensitivity = 65.2, specificity = 79.7), and 77.0% (sensitivity = 59.6, specificity = 86.1), respectively. Conclusions: Machine learning models trained for classification of CN versus AD can improve our prediction ability of MCI conversion to AD.

AB - Background: Predicting clinical course of cognitive decline can boost clinical trials' power and improve our clinical decision-making. Machine learning (ML) algorithms are specifically designed for the purpose of prediction; however. identifying optimal features or algorithms is still a challenge. Objective: To investigate the accuracy of different ML methods and different features to classify cognitively normal (CN) individuals from Alzheimer's disease (AD) and to predict longitudinal outcome in participants with mild cognitive impairment (MCI). Methods: A total of 1,329 participants from the Alzheimer's Disease Neuroimaging Initiative (ADNI) were included: 424 CN, 656 MCI, and 249 AD individuals. Four feature-sets at baseline (hippocampal volume and volume of 47 cortical and subcortical regions with and without demographics and APOE4) and six machine learning methods (decision trees, support vector machines, K-nearest neighbor, ensemble linear discriminant, boosted trees, and random forests) were used to classify participants with normal cognition from participants with AD. Subsequently the model with best classification performance was used for predicting clinical outcome of MCI participants. Results: Ensemble linear discriminant models using demographics and all volumetric magnetic resonance imaging measures as feature-set showed the best performance in classification of CN versus AD participants (accuracy = 92.8%, sensitivity = 95.8%, and specificity = 88.3%). Prediction accuracy of future conversion from MCI to AD for this ensemble linear discriminant at 6, 12, 24, 36, and 48 months was 63.8% (sensitivity = 74.4, specificity = 63.1), 68.9% (sensitivity = 75.9, specificity = 67.8), 74.9% (sensitivity = 71.5, specificity = 76.3), 75.3%, (sensitivity = 65.2, specificity = 79.7), and 77.0% (sensitivity = 59.6, specificity = 86.1), respectively. Conclusions: Machine learning models trained for classification of CN versus AD can improve our prediction ability of MCI conversion to AD.

KW - Alzheimer's disease

KW - classification

KW - early diagnosis

KW - machine learning

KW - magnetic resonance imaging

KW - mild cognitive impairment

KW - predictive analytics

UR - http://www.scopus.com/inward/record.url?scp=85073083277&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85073083277&partnerID=8YFLogxK

U2 - 10.3233/JAD-190262

DO - 10.3233/JAD-190262

M3 - Article

C2 - 31476152

AN - SCOPUS:85073083277

VL - 71

SP - 1027

EP - 1036

JO - Journal of Alzheimer's Disease

JF - Journal of Alzheimer's Disease

SN - 1387-2877

IS - 3

ER -