ROC Curves in Non‐Parametric Location‐Scale Regression Models

González–Manteiga, Wenceslao; Pardo–Fernández, Juan Carlos; Keilegom, Ingrid Van

doi:10.1111/j.1467-9469.2010.00693.x

Cited by 34 publications

(22 citation statements)

References 22 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…A common parametric approach to data analysis assumes that data from the D and

\overset{D}{true}

populations vary according to separate normal distributions. As a safeguard against model misspecification and to permit robustness from the sharp constraints of parametric models (e.g., the normal‐normal model) that can fail to accommodate increasingly complex distributions of data from modern medical tests, many contemporary methods for estimating test accuracy are based on flexible statistical models that use nonparametric or semiparametric structures, (e.g., Erkanli et al, ; Wang et al, ; Branscum et al, ; Hanson et al, ; Gonzalez‐Manteiga et al, ; Inácio et al, ; Inácio de Carvalho et al, ; Rodríguez and Martínez, ; Zhao et al, ). We develop a nonparametric Bayesian regression modeling framework that allows for data‐driven flexibility from the confines of parametric models by using dependent Dirichlet process (DDP) mixtures to estimate the covariate‐specific Youden index of a medical test and the covariate‐specific optimal threshold to screen individuals in practice.…”

Section: Introductionmentioning

confidence: 99%

Nonparametric Bayesian Covariate-Adjusted Estimation of the Youden Index

Carvalho

Branscum

2017

Biometrics

View full text Add to dashboard Cite

A novel nonparametric regression model is developed for evaluating the covariate-specific accuracy of a continuous biological marker. Accurately screening diseased from nondiseased individuals and correctly diagnosing disease stage are critically important to health care on several fronts, including guiding recommendations about combinations of treatments and their intensities. The accuracy of a continuous medical test or biomarker varies by the cutoff threshold (c) used to infer disease status. Accuracy can be measured by the probability of testing positive for diseased individuals (the true positive probability or sensitivity, Se(c), of the test), and the true negative probability (specificity, Sp(c)) of the test. A commonly used summary measure of test accuracy is the Youden index, YI=max{Se(c)+Sp(c)-1:c∈ℝ}, which is popular due in part to its ease of interpretation and relevance to population health research. In addition, clinical practitioners benefit from having an estimate of the optimal cutoff that maximizes sensitivity plus specificity available as a byproduct of estimating YI. We develop a highly flexible nonparametric model to estimate YI and its associated optimal cutoff that can respond to unanticipated skewness, multimodality, and other complexities because data distributions are modeled using dependent Dirichlet process mixtures. Important theoretical results on the support properties of the model are detailed. Inferences are available for the covariate-specific Youden index and its corresponding optimal cutoff threshold. The value of our nonparametric regression model is illustrated using multiple simulation studies and data on the age-specific accuracy of glucose as a biomarker of diabetes.

show abstract

“…A common parametric approach to data analysis assumes that data from the D and

\overset{D}{true}

Section: Introductionmentioning

confidence: 99%

Nonparametric Bayesian Covariate-Adjusted Estimation of the Youden Index

Carvalho

Branscum

2017

Biometrics

View full text Add to dashboard Cite

show abstract

“…Regarding the use of ROC curves, it should be taken into account that the classifiers that best discriminate between categories of state variables are those for which the curve is furthest away from the diagonal; this distance is indicated by the value of the “AUC” most different to 0.5. By representing the curves corresponding to the different classifying variables on the same graph, we can compare their efficacy and choose the variables with the most predictive power to obtain a correct classification (González‐Manteiga, Pardo‐Fernández & Van Keilegom, 2010).…”

Section: Resultsmentioning

confidence: 99%

The incorporation of ICT in higher education. The contribution of ROC curves in the graphic visualization of differences in the analysis of the variables

Muñoz-Repiso

Tejedor

2012

Brit J Educational Tech

View full text Add to dashboard Cite

This paper has focused on the productive use of information and communications technology (ICT) by university students and its influence on academic performance. The objective is to determine whether a comparison of successful and non-successful students (in relation to academic achievements) versus the variables studied (linked to the process of ICT incorporation) can help us to establish contextual and behavioral patterns associated with good academic performance. This differential analysis is carried out by making basic statistical comparisons (student's t-test). The data analysis is completed by applying receiver operating characteristic curves, a methodology of new application in the area of education. The sample was composed of 1194 students enrolled in 40 different subjects at various universities in two countries, Mexico and Spain. Students gave a very positive appraisal of the value of ICTs to support the teaching process, but they express an intermediate level of satisfaction with the teaching conditions in subjects, which are using the support of these technologies. Moreover, there is a clear relation between high scores in the variables associated with the processes of incorporation of ICT (satisfaction with teaching, higher levels of competence in ICT, less training needs in ICT, more favorable attitudes toward their integration in training processes) and achieving higher academic qualifications.which enable them to realize the full potential of the technology (

show abstract

“…Many of these models have been programmed in the suite of R functions, DP Package: Bayesian Nonparametric Modeling in R, which can be found at http://cran.r-project.org/web/packages/DPpackage/index.html. Nonparametric frequentist methods for gold-standard ROC analysis using location-scale models are also available [64]. * contain a first column of all ones to accommodate intercepts.…”

Section: Discussionmentioning

confidence: 99%

Flexible regression models for ROC and risk analysis, with or without a gold standard

Branscum

Johnson

Hanson

et al. 2015

Statistics in Medicine

View full text Add to dashboard Cite

A novel semiparametric regression model is developed for evaluating the covariate-specific accuracy of a continuous medical test or biomarker. Ideally, studies designed to estimate or compare medical test accuracy will use a separate, flawless gold-standard procedure to determine the true disease status of sampled individuals. We treat this as a special case of the more complicated and increasingly common scenario in which disease status is unknown because a gold-standard procedure does not exist or is too costly or invasive for widespread use. To compensate for missing data on disease status, covariate information is used to discriminate between diseased and healthy units. We thus model the probability of disease as a function of 'disease covariates'. In addition, we model test/biomarker outcome data to depend on 'test covariates', which provides researchers the opportunity to quantify the impact of covariates on the accuracy of a medical test. We further model the distributions of test outcomes using flexible semiparametric classes. An important new theoretical result demonstrating model identifiability under mild conditions is presented. The modeling framework can be used to obtain inferences about covariate-specific test accuracy and the probability of disease based on subject-specific disease and test covariate information. The value of the model is illustrated using multiple simulation studies and data on the age-adjusted ability of soluble epidermal growth factor receptor -a ubiquitous serum protein -to serve as a biomarker of lung cancer in men. sas code for fitting the model is provided.

show abstract

ROC Curves in Non‐Parametric Location‐Scale Regression Models

Cited by 34 publications

References 22 publications

Nonparametric Bayesian Covariate-Adjusted Estimation of the Youden Index

Nonparametric Bayesian Covariate-Adjusted Estimation of the Youden Index

The incorporation of ICT in higher education. The contribution of ROC curves in the graphic visualization of differences in the analysis of the variables

Flexible regression models for ROC and risk analysis, with or without a gold standard

Contact Info

Product

Resources

About