Functional covariate-adjusted partial area under the specificity-ROC curve with an application to metabolic syndrome diagnosis

Inácio, Vanda; Carvalho, M.; Alonzo, Todd A.; González–Manteiga, Wenceslao

doi:10.1214/16-aoas943

Cited by 14 publications

(11 citation statements)

References 24 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Discrimination surfaces—as defined above—have connections with the area under conditional ROC curves, defined as

AUC false(bold-italicx false) = \int_{0}^{1} ROC false(p false| bold-italicx false) normald p

, where

ROC false(p false| bold-italicx false) = 1 - F_{D} false(F_{\overset{D}{true}}^{- 1} false(1 - p false| bold-italicx false) false| bold-italicx false)

, with x in

R^{p}

being a covariate. Yet as it can be seen from Equation (and as it will be seen from Equation ), here, the target is on seeking for regions of maximum discrimination and not simply on assessing how discrimination ability changes over a covariate.…”

Section: Discrimination Surfacesmentioning

confidence: 99%

“…Such surfaces and corresponding sets allow us to identify regions of the brain at which left‐to‐right morphological differences between diseased and nondiseased participants are more likely to occur. From a conceptual viewpoint, discrimination surfaces—as formally defined in Section 3—have connections with the area under conditional ROC curves, as discussed for example by Inácio de Carvalho et al However, while in conditional ROC curves, the objective is often on assessing how the discrimination ability of a diagnostic test changes over a predictor; here, the goal is on searching for regions of maximum discrimination. Another object, which has connections with discrimination surfaces, is the so‐called free‐response receiver operating characteristic (ROC) curve; yet an important distinction between the 2 paradigms is that discrimination surfaces deliver as output a suspected location, whereas free‐response ROC curve take as input suspected locations—along with a rating on the level of suspicion.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Discrimination surfaces with application to region‐specific brain asymmetry analysis

Martos

Carvalho

2018

Statistics in Medicine

Self Cite

View full text Add to dashboard Cite

Discrimination surfaces are here introduced as a diagnostic tool for localizing brain regions where discrimination between diseased and nondiseased participants is higher. To estimate discrimination surfaces, we introduce a Mann-Whitney type of statistic for random fields and present large-sample results characterizing its asymptotic behavior. Simulation results demonstrate that our estimator accurately recovers the true surface and corresponding interval of maximal discrimination. The empirical analysis suggests that in the anterior region of the brain, schizophrenic patients tend to present lower local asymmetry scores in comparison with participants in the control group.

show abstract

“…Discrimination surfaces—as defined above—have connections with the area under conditional ROC curves, defined as

AUC false(bold-italicx false) = \int_{0}^{1} ROC false(p false| bold-italicx false) normald p

, where

ROC false(p false| bold-italicx false) = 1 - F_{D} false(F_{\overset{D}{true}}^{- 1} false(1 - p false| bold-italicx false) false| bold-italicx false)

, with x in

R^{p}

Section: Discrimination Surfacesmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Discrimination surfaces with application to region‐specific brain asymmetry analysis

Martos

Carvalho

2018

Statistics in Medicine

Self Cite

View full text Add to dashboard Cite

show abstract

“…This method has been recently generalised to functional covariates by Inàcio de Carvalho et al. 30 In that paper, the authors propose a functional conditional partial area under the specificity ROC curve. The estimator by Yao et al.…”

Section: Modelling Covariate Effects On the Roc Curvementioning

confidence: 99%

“…In a fully nonparametric setting, Yao et al 16 present a 'conditional' Mann-Whitney estimator for AUC x estimation. This method has been recently generalised to functional covariates by Ina`cio de Carvalho et al 30 In that paper, the authors propose a functional conditional partial area under the specificity ROC curve. The estimator by Yao et al 16 is a particular case when the sensitivity is not restricted to a specific interval.…”

Section: Area Under the Conditional Roc Curvementioning

confidence: 99%

Bootstrap-based procedures for inference in nonparametric receiver-operating characteristic curve regression analysis

Rodríguez‐Álvarez

Roca‐Pardiñas

Cadarso‐Suárez

et al. 2017

Stat Methods Med Res

View full text Add to dashboard Cite

This document contains supplementary material to the paper "Bootstrap-based procedures for inference in nonparametric receiver-operating characteristic curve regression analysis". Additional simulation studies to complete those presented in the main manuscript are provided. More precisely, we present here the results when considering different distributions for simulating the diagnostic test result in healthy and diseased populations, and when covariates only affect the result of the diagnostic test in the diseased population. Web Appendix A Simulation study with different distributionsData were simulated from two scenarios, namely,• Scenario wI

show abstract

“…Through the Karhunen‐Loéve orthogonal expansion, Müller and Stadtmüller represented a random predictor function by the first several FPC scores and approximated a generalized functional linear regression model by a generalized linear regression model in which the selected FPC scores were used as covariates in the model. The FDA approach was used by Inácio et al in developing a functional covariate adjusted estimator for the area under the ROC curve (AUC) and the partial AUC.…”

Section: Introductionmentioning

confidence: 99%

Combining biomarker trajectories to improve diagnostic accuracy in prospective cohort studies with verification bias

Gatsonis

2018

Statistics in Medicine

View full text Add to dashboard Cite

In this paper, we develop methods to combine multiple biomarker trajectories into a composite diagnostic marker using functional data analysis (FDA) to achieve better diagnostic accuracy in monitoring disease recurrence in the setting of a prospective cohort study. In such studies, the disease status is usually verified only for patients with a positive test result in any biomarker and is missing in patients with negative test results in all biomarkers. Thus, the test result will affect disease verification, which leads to verification bias if the analysis is restricted only to the verified cases. We treat verification bias as a missing data problem. Under both missing at random (MAR) and missing not at random (MNAR) assumptions, we derive the optimal classification rules using the Neyman‐Pearson lemma based on the composite diagnostic marker. We estimate thresholds adjusted for verification bias to dichotomize patients as test positive or test negative, and we evaluate the diagnostic accuracy using the verification bias corrected area under the ROC curves (AUCs). We evaluate the performance and robustness of the FDA combination approach and assess the consistency of the approach through simulation studies. In addition, we perform a sensitivity analysis of the dependency between the verification process and disease status for the approach under the MNAR assumption. We apply the proposed method on data from the Religious Orders Study and from a non‐small cell lung cancer trial.

show abstract

Functional covariate-adjusted partial area under the specificity-ROC curve with an application to metabolic syndrome diagnosis

Cited by 14 publications

References 24 publications

Discrimination surfaces with application to region‐specific brain asymmetry analysis

Discrimination surfaces with application to region‐specific brain asymmetry analysis

Bootstrap-based procedures for inference in nonparametric receiver-operating characteristic curve regression analysis

Combining biomarker trajectories to improve diagnostic accuracy in prospective cohort studies with verification bias

Contact Info

Product

Resources

About