Estimation of ROC curves based on stably distributed biomarkers subject to measurement error and pooling mixtures

Albert Vexler, Enrique F. Schisterman, Aiyi Liu

Research output: Contribution to journalArticle

Abstract

Additive measurement errors and pooling design are objectively two different issues, which have been separately and extensively dealt with in the biostatistics literature. However, these topics usually correspond to problems of reconstructing a summand's distribution of the biomarker by the distribution of the convoluted observations. Thus, we associate the two issues into one stated problem. The integrated approach creates an opportunity to investigate new fields, e.g. a subject of pooling errors, issues regarding pooled data affected by measurement errors. To be specific, we consider the stated problem in the context of the receiver operating characteristic (ROC) curves analysis, which is the well-accepted tool for evaluating the ability of a biomarker to discriminate between two populations. The present paper considers a wide family of biospecimen distributions. In addition, applied assumptions, which are related to distribution functions of biomarkers, are mainly conditioned by the reconstructing problem. We propose and examine maximum likelihood techniques based on the following data: a biomarker with measurement error; pooled samples; and pooled samples with measurement error. The obtained methods are illustrated by applications to real data studies.

Original languageEnglish (US)
Pages (from-to)280-296
Number of pages17
JournalStatistics in Medicine
Volume27
Issue number2
DOIs
StatePublished - Jan 30 2008
Externally publishedYes

Fingerprint

Pooling
Receiver Operating Characteristic Curve
Biomarkers
Measurement Error
ROC Curve
Pooling Designs
Biostatistics
Maximum Likelihood
Distribution Function
Population

Keywords

  • Deconvolution
  • Design of experiments
  • Fourier inversion
  • Infinitely divisible distribution
  • Measurement error
  • Pooling blood samples
  • Receiver operating characteristic curves
  • Stable distribution
  • Summand's distribution

ASJC Scopus subject areas

  • Epidemiology

Cite this

Estimation of ROC curves based on stably distributed biomarkers subject to measurement error and pooling mixtures. / Vexler, Albert; Schisterman, Enrique F.; Liu, Aiyi.

In: Statistics in Medicine, Vol. 27, No. 2, 30.01.2008, p. 280-296.

Research output: Contribution to journalArticle

Vexler, Albert ; Schisterman, Enrique F. ; Liu, Aiyi. / Estimation of ROC curves based on stably distributed biomarkers subject to measurement error and pooling mixtures. In: Statistics in Medicine. 2008 ; Vol. 27, No. 2. pp. 280-296.
@article{9c6fa49e401043eebcea3c07b0062973,
title = "Estimation of ROC curves based on stably distributed biomarkers subject to measurement error and pooling mixtures",
abstract = "Additive measurement errors and pooling design are objectively two different issues, which have been separately and extensively dealt with in the biostatistics literature. However, these topics usually correspond to problems of reconstructing a summand's distribution of the biomarker by the distribution of the convoluted observations. Thus, we associate the two issues into one stated problem. The integrated approach creates an opportunity to investigate new fields, e.g. a subject of pooling errors, issues regarding pooled data affected by measurement errors. To be specific, we consider the stated problem in the context of the receiver operating characteristic (ROC) curves analysis, which is the well-accepted tool for evaluating the ability of a biomarker to discriminate between two populations. The present paper considers a wide family of biospecimen distributions. In addition, applied assumptions, which are related to distribution functions of biomarkers, are mainly conditioned by the reconstructing problem. We propose and examine maximum likelihood techniques based on the following data: a biomarker with measurement error; pooled samples; and pooled samples with measurement error. The obtained methods are illustrated by applications to real data studies.",
keywords = "Deconvolution, Design of experiments, Fourier inversion, Infinitely divisible distribution, Measurement error, Pooling blood samples, Receiver operating characteristic curves, Stable distribution, Summand's distribution",
author = "Albert Vexler and Schisterman, {Enrique F.} and Aiyi Liu",
year = "2008",
month = "1",
day = "30",
doi = "10.1002/sim.3035",
language = "English (US)",
volume = "27",
pages = "280--296",
journal = "Statistics in Medicine",
issn = "0277-6715",
publisher = "John Wiley and Sons Ltd",
number = "2",

}

TY - JOUR

T1 - Estimation of ROC curves based on stably distributed biomarkers subject to measurement error and pooling mixtures

AU - Vexler, Albert

AU - Schisterman, Enrique F.

AU - Liu, Aiyi

PY - 2008/1/30

Y1 - 2008/1/30

N2 - Additive measurement errors and pooling design are objectively two different issues, which have been separately and extensively dealt with in the biostatistics literature. However, these topics usually correspond to problems of reconstructing a summand's distribution of the biomarker by the distribution of the convoluted observations. Thus, we associate the two issues into one stated problem. The integrated approach creates an opportunity to investigate new fields, e.g. a subject of pooling errors, issues regarding pooled data affected by measurement errors. To be specific, we consider the stated problem in the context of the receiver operating characteristic (ROC) curves analysis, which is the well-accepted tool for evaluating the ability of a biomarker to discriminate between two populations. The present paper considers a wide family of biospecimen distributions. In addition, applied assumptions, which are related to distribution functions of biomarkers, are mainly conditioned by the reconstructing problem. We propose and examine maximum likelihood techniques based on the following data: a biomarker with measurement error; pooled samples; and pooled samples with measurement error. The obtained methods are illustrated by applications to real data studies.

AB - Additive measurement errors and pooling design are objectively two different issues, which have been separately and extensively dealt with in the biostatistics literature. However, these topics usually correspond to problems of reconstructing a summand's distribution of the biomarker by the distribution of the convoluted observations. Thus, we associate the two issues into one stated problem. The integrated approach creates an opportunity to investigate new fields, e.g. a subject of pooling errors, issues regarding pooled data affected by measurement errors. To be specific, we consider the stated problem in the context of the receiver operating characteristic (ROC) curves analysis, which is the well-accepted tool for evaluating the ability of a biomarker to discriminate between two populations. The present paper considers a wide family of biospecimen distributions. In addition, applied assumptions, which are related to distribution functions of biomarkers, are mainly conditioned by the reconstructing problem. We propose and examine maximum likelihood techniques based on the following data: a biomarker with measurement error; pooled samples; and pooled samples with measurement error. The obtained methods are illustrated by applications to real data studies.

KW - Deconvolution

KW - Design of experiments

KW - Fourier inversion

KW - Infinitely divisible distribution

KW - Measurement error

KW - Pooling blood samples

KW - Receiver operating characteristic curves

KW - Stable distribution

KW - Summand's distribution

UR - http://www.scopus.com/inward/record.url?scp=38849182242&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=38849182242&partnerID=8YFLogxK

U2 - 10.1002/sim.3035

DO - 10.1002/sim.3035

M3 - Article

C2 - 17721905

AN - SCOPUS:38849182242

VL - 27

SP - 280

EP - 296

JO - Statistics in Medicine

JF - Statistics in Medicine

SN - 0277-6715

IS - 2

ER -