Maximizing power to track Alzheimer's disease and MCI progression by LDA-based weighting of longitudinal ventricular surface features

Boris A. Gutman, Xue Hua, Priya Rajagopalan, Yi Yu Chou, Yalin Wang, Igor Yanovsky, Arthur W. Toga, Clifford R Jr. Jack, Michael W. Weiner, Paul M. Thompson

Research output: Contribution to journalArticle

39 Citations (Scopus)

Abstract

We propose a new method to maximize biomarker efficiency for detecting anatomical change over time in serial MRI. Drug trials using neuroimaging become prohibitively costly if vast numbers of subjects must be assessed, so it is vital to develop efficient measures of brain change. A popular measure of efficiency is the minimal sample size (n80) needed to detect 25% change in a biomarker, with 95% confidence and 80% power. For multivariate measures of brain change, we can directly optimize n80 based on a Linear Discriminant Analysis (LDA). Here we use a supervised learning framework to optimize n80, offering two alternative solutions. With a new medial surface modeling method, we track 3D dynamic changes in the lateral ventricles in 2065 ADNI scans. We apply our LDA-based weighting to the results. Our best average n80-in two-fold nested cross-validation-is 104 MCI subjects (95% CI: [94,139]) for a 1-year drug trial, and 75. AD subjects [64,102]. This compares favorably with other MRI analysis methods. The standard "statistical ROI" approach applied to the same ventricular surfaces requires 165 MCI or 94. AD subjects. At 2. years, the best LDA measure needs only 67 MCI and 52. AD subjects, versus 119 MCI and 80. AD subjects for the stat-ROI method. Our surface-based measures are unbiased: they give no artifactual additive atrophy over three time points. Our results suggest that statistical weighting may boost efficiency of drug trials that use brain maps.

Original languageEnglish (US)
Pages (from-to)386-401
Number of pages16
JournalNeuroImage
Volume70
DOIs
StatePublished - Apr 15 2013

Fingerprint

Discriminant Analysis
Disease Progression
Alzheimer Disease
Efficiency
Brain
Biomarkers
Pharmaceutical Preparations
Lateral Ventricles
Neuroimaging
Sample Size
Atrophy
Learning
Power (Psychology)

Keywords

  • ADNI
  • Alzheimer's disease
  • Biomarker
  • Drug trial
  • Lateral ventricles
  • Linear Discriminant Analysis
  • Machine learning
  • Mild cognitive impairment
  • Shape analysis

ASJC Scopus subject areas

  • Cognitive Neuroscience
  • Neurology

Cite this

Maximizing power to track Alzheimer's disease and MCI progression by LDA-based weighting of longitudinal ventricular surface features. / Gutman, Boris A.; Hua, Xue; Rajagopalan, Priya; Chou, Yi Yu; Wang, Yalin; Yanovsky, Igor; Toga, Arthur W.; Jack, Clifford R Jr.; Weiner, Michael W.; Thompson, Paul M.

In: NeuroImage, Vol. 70, 15.04.2013, p. 386-401.

Research output: Contribution to journalArticle

Gutman, BA, Hua, X, Rajagopalan, P, Chou, YY, Wang, Y, Yanovsky, I, Toga, AW, Jack, CRJ, Weiner, MW & Thompson, PM 2013, 'Maximizing power to track Alzheimer's disease and MCI progression by LDA-based weighting of longitudinal ventricular surface features', NeuroImage, vol. 70, pp. 386-401. https://doi.org/10.1016/j.neuroimage.2012.12.052
Gutman, Boris A. ; Hua, Xue ; Rajagopalan, Priya ; Chou, Yi Yu ; Wang, Yalin ; Yanovsky, Igor ; Toga, Arthur W. ; Jack, Clifford R Jr. ; Weiner, Michael W. ; Thompson, Paul M. / Maximizing power to track Alzheimer's disease and MCI progression by LDA-based weighting of longitudinal ventricular surface features. In: NeuroImage. 2013 ; Vol. 70. pp. 386-401.
@article{11b9079abaca48fe9062c893b15a487f,
title = "Maximizing power to track Alzheimer's disease and MCI progression by LDA-based weighting of longitudinal ventricular surface features",
abstract = "We propose a new method to maximize biomarker efficiency for detecting anatomical change over time in serial MRI. Drug trials using neuroimaging become prohibitively costly if vast numbers of subjects must be assessed, so it is vital to develop efficient measures of brain change. A popular measure of efficiency is the minimal sample size (n80) needed to detect 25{\%} change in a biomarker, with 95{\%} confidence and 80{\%} power. For multivariate measures of brain change, we can directly optimize n80 based on a Linear Discriminant Analysis (LDA). Here we use a supervised learning framework to optimize n80, offering two alternative solutions. With a new medial surface modeling method, we track 3D dynamic changes in the lateral ventricles in 2065 ADNI scans. We apply our LDA-based weighting to the results. Our best average n80-in two-fold nested cross-validation-is 104 MCI subjects (95{\%} CI: [94,139]) for a 1-year drug trial, and 75. AD subjects [64,102]. This compares favorably with other MRI analysis methods. The standard {"}statistical ROI{"} approach applied to the same ventricular surfaces requires 165 MCI or 94. AD subjects. At 2. years, the best LDA measure needs only 67 MCI and 52. AD subjects, versus 119 MCI and 80. AD subjects for the stat-ROI method. Our surface-based measures are unbiased: they give no artifactual additive atrophy over three time points. Our results suggest that statistical weighting may boost efficiency of drug trials that use brain maps.",
keywords = "ADNI, Alzheimer's disease, Biomarker, Drug trial, Lateral ventricles, Linear Discriminant Analysis, Machine learning, Mild cognitive impairment, Shape analysis",
author = "Gutman, {Boris A.} and Xue Hua and Priya Rajagopalan and Chou, {Yi Yu} and Yalin Wang and Igor Yanovsky and Toga, {Arthur W.} and Jack, {Clifford R Jr.} and Weiner, {Michael W.} and Thompson, {Paul M.}",
year = "2013",
month = "4",
day = "15",
doi = "10.1016/j.neuroimage.2012.12.052",
language = "English (US)",
volume = "70",
pages = "386--401",
journal = "NeuroImage",
issn = "1053-8119",
publisher = "Academic Press Inc.",

}

TY - JOUR

T1 - Maximizing power to track Alzheimer's disease and MCI progression by LDA-based weighting of longitudinal ventricular surface features

AU - Gutman, Boris A.

AU - Hua, Xue

AU - Rajagopalan, Priya

AU - Chou, Yi Yu

AU - Wang, Yalin

AU - Yanovsky, Igor

AU - Toga, Arthur W.

AU - Jack, Clifford R Jr.

AU - Weiner, Michael W.

AU - Thompson, Paul M.

PY - 2013/4/15

Y1 - 2013/4/15

N2 - We propose a new method to maximize biomarker efficiency for detecting anatomical change over time in serial MRI. Drug trials using neuroimaging become prohibitively costly if vast numbers of subjects must be assessed, so it is vital to develop efficient measures of brain change. A popular measure of efficiency is the minimal sample size (n80) needed to detect 25% change in a biomarker, with 95% confidence and 80% power. For multivariate measures of brain change, we can directly optimize n80 based on a Linear Discriminant Analysis (LDA). Here we use a supervised learning framework to optimize n80, offering two alternative solutions. With a new medial surface modeling method, we track 3D dynamic changes in the lateral ventricles in 2065 ADNI scans. We apply our LDA-based weighting to the results. Our best average n80-in two-fold nested cross-validation-is 104 MCI subjects (95% CI: [94,139]) for a 1-year drug trial, and 75. AD subjects [64,102]. This compares favorably with other MRI analysis methods. The standard "statistical ROI" approach applied to the same ventricular surfaces requires 165 MCI or 94. AD subjects. At 2. years, the best LDA measure needs only 67 MCI and 52. AD subjects, versus 119 MCI and 80. AD subjects for the stat-ROI method. Our surface-based measures are unbiased: they give no artifactual additive atrophy over three time points. Our results suggest that statistical weighting may boost efficiency of drug trials that use brain maps.

AB - We propose a new method to maximize biomarker efficiency for detecting anatomical change over time in serial MRI. Drug trials using neuroimaging become prohibitively costly if vast numbers of subjects must be assessed, so it is vital to develop efficient measures of brain change. A popular measure of efficiency is the minimal sample size (n80) needed to detect 25% change in a biomarker, with 95% confidence and 80% power. For multivariate measures of brain change, we can directly optimize n80 based on a Linear Discriminant Analysis (LDA). Here we use a supervised learning framework to optimize n80, offering two alternative solutions. With a new medial surface modeling method, we track 3D dynamic changes in the lateral ventricles in 2065 ADNI scans. We apply our LDA-based weighting to the results. Our best average n80-in two-fold nested cross-validation-is 104 MCI subjects (95% CI: [94,139]) for a 1-year drug trial, and 75. AD subjects [64,102]. This compares favorably with other MRI analysis methods. The standard "statistical ROI" approach applied to the same ventricular surfaces requires 165 MCI or 94. AD subjects. At 2. years, the best LDA measure needs only 67 MCI and 52. AD subjects, versus 119 MCI and 80. AD subjects for the stat-ROI method. Our surface-based measures are unbiased: they give no artifactual additive atrophy over three time points. Our results suggest that statistical weighting may boost efficiency of drug trials that use brain maps.

KW - ADNI

KW - Alzheimer's disease

KW - Biomarker

KW - Drug trial

KW - Lateral ventricles

KW - Linear Discriminant Analysis

KW - Machine learning

KW - Mild cognitive impairment

KW - Shape analysis

UR - http://www.scopus.com/inward/record.url?scp=84873807104&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84873807104&partnerID=8YFLogxK

U2 - 10.1016/j.neuroimage.2012.12.052

DO - 10.1016/j.neuroimage.2012.12.052

M3 - Article

VL - 70

SP - 386

EP - 401

JO - NeuroImage

JF - NeuroImage

SN - 1053-8119

ER -