Principal investigator: Aiyi Liu, Ph.D.
This research focuses on developing statistical methods for biomarkers to address various issues arising from studies assessing the association of biomarkers and certain diseases or health conditions. These issues include evaluation and comparison of diagnostic accuracy and measurement errors of biomarkers, statistically sound analysis of data that are subject to a limit of detection, and pooling biospecimens. For diagnostic accuracy, research concentrates on inference of the Receiver Operating Characteristic (ROC) curves with various types of data, especially data from repeated measurements, data from pooled biospecimens, and/or data subject to a limit of detection, and on combination of biomarkers to improve diagnostic accuracy. Developing multi-stage procedures to evaluate and compare the accuracy of biomarkers is another focus of this effort. For assessing measurement errors, research focuses on developing multi-stage evaluation procedures and on designing analyses following these procedures. For studies involving limit of detection of a biomarker, research focuses on assessing the effects of the data below the detection-limit measurement errors on various inferential objectives.
DESPR Collaborators
·
Enrique F. Schisterman, Ph.D. ·
Kai F. Yu, Ph.D. Selected Publications
Vexler A, Liu A, Eliseeva E, & Schisterman EF. (In press). Maximum likelihood ratio tests for comparing the discriminatory ability of biomarkers subject to limit of detection. Biometrics.
Liu A, Wu CQ, & Yu KF. (2008). Two-stage procedures for selecting the best diagnostic biomarkers. Philosophical Transaction of the Royal Society A, 366, 2293-2299.
Liu A, Wu CQ, & Schisterman EF. (2008). Nonparametric sequential evaluation of diagnostic biomarkers. Statistics in Medicine, 27(10):1667-1678. [Abstract]
Liu A, Schisterman EF, & Wu CQ. (2006). Multistage evaluation of measurement error in a reliability study. Biometrics, 62(4):1190-1196. [Abstract]