Improved nonparametric estimation of the optimal diagnostic cut-off point associated with the Youden index under different sampling schemes

Jingjing Yin, Hani Samawi, Daniel Linder

Research output: Contribution to journalArticle

9 Citations (Scopus)

Abstract

A diagnostic cut-off point of a biomarker measurement is needed for classifying a random subject to be either diseased or healthy. However, the cut-off point is usually unknown and needs to be estimated by some optimization criteria. One important criterion is the Youden index, which has been widely adopted in practice. The Youden index, which is defined as the maximum of (sensitivity + specificity −1), directly measures the largest total diagnostic accuracy a biomarker can achieve. Therefore, it is desirable to estimate the optimal cut-off point associated with the Youden index. Sometimes, taking the actual measurements of a biomarker is very difficult and expensive, while ranking them without the actual measurement can be relatively easy. In such cases, ranked set sampling can give more precise estimation than simple random sampling, as ranked set samples are more likely to span the full range of the population. In this study, kernel density estimation is utilized to numerically solve for an estimate of the optimal cut-off point. The asymptotic distributions of the kernel estimators based on two sampling schemes are derived analytically and we prove that the estimators based on ranked set sampling are relatively more efficient than that of simple random sampling and both estimators are asymptotically unbiased. Furthermore, the asymptotic confidence intervals are derived. Intensive simulations are carried out to compare the proposed method using ranked set sampling with simple random sampling, with the proposed method outperforming simple random sampling in all cases. A real data set is analyzed for illustrating the proposed method.

Original languageEnglish (US)
Pages (from-to)915-934
Number of pages20
JournalBiometrical Journal
Volume58
Issue number4
DOIs
StatePublished - Jul 1 2016

Fingerprint

Simple Random Sampling
Nonparametric Estimation
Ranked Set Sampling
Diagnostics
Biomarkers
Ranked Set Sample
Diagnostic Accuracy
Estimator
Kernel Density Estimation
Kernel Estimator
Estimate
Asymptotic distribution
Specificity
Confidence interval
Ranking
Likely
Unknown
Sampling
Nonparametric estimation
Random sampling

Keywords

  • Diagnostic accuracy
  • Kernel smoothing
  • ROC analysis
  • Ranked set sampling
  • Youden index

ASJC Scopus subject areas

  • Statistics and Probability
  • Statistics, Probability and Uncertainty

Cite this

Improved nonparametric estimation of the optimal diagnostic cut-off point associated with the Youden index under different sampling schemes. / Yin, Jingjing; Samawi, Hani; Linder, Daniel.

In: Biometrical Journal, Vol. 58, No. 4, 01.07.2016, p. 915-934.

Research output: Contribution to journalArticle

@article{42370be12f494db8905c3d5a8d59db1f,
title = "Improved nonparametric estimation of the optimal diagnostic cut-off point associated with the Youden index under different sampling schemes",
abstract = "A diagnostic cut-off point of a biomarker measurement is needed for classifying a random subject to be either diseased or healthy. However, the cut-off point is usually unknown and needs to be estimated by some optimization criteria. One important criterion is the Youden index, which has been widely adopted in practice. The Youden index, which is defined as the maximum of (sensitivity + specificity −1), directly measures the largest total diagnostic accuracy a biomarker can achieve. Therefore, it is desirable to estimate the optimal cut-off point associated with the Youden index. Sometimes, taking the actual measurements of a biomarker is very difficult and expensive, while ranking them without the actual measurement can be relatively easy. In such cases, ranked set sampling can give more precise estimation than simple random sampling, as ranked set samples are more likely to span the full range of the population. In this study, kernel density estimation is utilized to numerically solve for an estimate of the optimal cut-off point. The asymptotic distributions of the kernel estimators based on two sampling schemes are derived analytically and we prove that the estimators based on ranked set sampling are relatively more efficient than that of simple random sampling and both estimators are asymptotically unbiased. Furthermore, the asymptotic confidence intervals are derived. Intensive simulations are carried out to compare the proposed method using ranked set sampling with simple random sampling, with the proposed method outperforming simple random sampling in all cases. A real data set is analyzed for illustrating the proposed method.",
keywords = "Diagnostic accuracy, Kernel smoothing, ROC analysis, Ranked set sampling, Youden index",
author = "Jingjing Yin and Hani Samawi and Daniel Linder",
year = "2016",
month = "7",
day = "1",
doi = "10.1002/bimj.201500036",
language = "English (US)",
volume = "58",
pages = "915--934",
journal = "Biometrical Journal",
issn = "0323-3847",
publisher = "Wiley-VCH Verlag",
number = "4",

}

TY - JOUR

T1 - Improved nonparametric estimation of the optimal diagnostic cut-off point associated with the Youden index under different sampling schemes

AU - Yin, Jingjing

AU - Samawi, Hani

AU - Linder, Daniel

PY - 2016/7/1

Y1 - 2016/7/1

N2 - A diagnostic cut-off point of a biomarker measurement is needed for classifying a random subject to be either diseased or healthy. However, the cut-off point is usually unknown and needs to be estimated by some optimization criteria. One important criterion is the Youden index, which has been widely adopted in practice. The Youden index, which is defined as the maximum of (sensitivity + specificity −1), directly measures the largest total diagnostic accuracy a biomarker can achieve. Therefore, it is desirable to estimate the optimal cut-off point associated with the Youden index. Sometimes, taking the actual measurements of a biomarker is very difficult and expensive, while ranking them without the actual measurement can be relatively easy. In such cases, ranked set sampling can give more precise estimation than simple random sampling, as ranked set samples are more likely to span the full range of the population. In this study, kernel density estimation is utilized to numerically solve for an estimate of the optimal cut-off point. The asymptotic distributions of the kernel estimators based on two sampling schemes are derived analytically and we prove that the estimators based on ranked set sampling are relatively more efficient than that of simple random sampling and both estimators are asymptotically unbiased. Furthermore, the asymptotic confidence intervals are derived. Intensive simulations are carried out to compare the proposed method using ranked set sampling with simple random sampling, with the proposed method outperforming simple random sampling in all cases. A real data set is analyzed for illustrating the proposed method.

AB - A diagnostic cut-off point of a biomarker measurement is needed for classifying a random subject to be either diseased or healthy. However, the cut-off point is usually unknown and needs to be estimated by some optimization criteria. One important criterion is the Youden index, which has been widely adopted in practice. The Youden index, which is defined as the maximum of (sensitivity + specificity −1), directly measures the largest total diagnostic accuracy a biomarker can achieve. Therefore, it is desirable to estimate the optimal cut-off point associated with the Youden index. Sometimes, taking the actual measurements of a biomarker is very difficult and expensive, while ranking them without the actual measurement can be relatively easy. In such cases, ranked set sampling can give more precise estimation than simple random sampling, as ranked set samples are more likely to span the full range of the population. In this study, kernel density estimation is utilized to numerically solve for an estimate of the optimal cut-off point. The asymptotic distributions of the kernel estimators based on two sampling schemes are derived analytically and we prove that the estimators based on ranked set sampling are relatively more efficient than that of simple random sampling and both estimators are asymptotically unbiased. Furthermore, the asymptotic confidence intervals are derived. Intensive simulations are carried out to compare the proposed method using ranked set sampling with simple random sampling, with the proposed method outperforming simple random sampling in all cases. A real data set is analyzed for illustrating the proposed method.

KW - Diagnostic accuracy

KW - Kernel smoothing

KW - ROC analysis

KW - Ranked set sampling

KW - Youden index

UR - http://www.scopus.com/inward/record.url?scp=85027941411&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85027941411&partnerID=8YFLogxK

U2 - 10.1002/bimj.201500036

DO - 10.1002/bimj.201500036

M3 - Article

C2 - 26756282

AN - SCOPUS:85027941411

VL - 58

SP - 915

EP - 934

JO - Biometrical Journal

JF - Biometrical Journal

SN - 0323-3847

IS - 4

ER -