|
Activity Number:
|
246
- Improved Disease Classification Through Extensions of ROC Curve Estimation and Biomarker Characterization
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Tuesday, August 9, 2022 : 8:30 AM to 10:20 AM
|
|
Sponsor:
|
Section on Medical Devices and Diagnostics
|
|
Abstract #322658
|
|
|
Title:
|
Parametric and Semiparametric Multiple Imputation for Missing Biomarker Values
|
|
Author(s):
|
Peng Shi* and Leonidas Bantis
|
|
Companies:
|
University of Kansas Medical Center and University of Kansas Medical Center
|
|
Keywords:
|
AFT;
Biomarker;
HCNS;
Multiple Imputation;
ROC
|
|
Abstract:
|
The ROC curve is a well-known statistical tool for assessing the biomarkers’ discriminatory ability. In practice, biomarker data of a given study could suffer from missing values. In such situations, one could conduct the ROC analysis based only on the complete cases. However, this will come at a cost of efficiency and potentially lead to biased estimates regarding the accuracy of the marker. In this work, we study and propose a multiple imputation framework that operates parametrically or semi-parametrically. It involves the use of accelerated failure time (AFT) models and a hazard constrained natural spline (HCNS) approach. We evaluate our approaches through extensive simulations and illustrate the proposed methods using a real data set.
|
Authors who are presenting talks have a * after their name.
