A1672
Title: Accelerated failure time models with error-prone response and nonlinear covariates
Authors: Li-Pang Chen - National Chengchi University (Taiwan) [presenting]
Abstract: Survival analysis of patient survival time for a specific Cancer using gene expressions as covariates is a main interest in medical studies. In the framework of survival analysis, the accelerated failure time model in parametric form is a common approach. However, gene expressions are possibly nonlinear and the survival time as well as censoring status are subject to measurement error. To address these complex features simultaneously, measurement error in survival time and censoring status is first corrected, and these corrections are used to develop a corrected Buckley-James estimator. Subsequently, the boosting algorithm with cubic spline estimation is used to iteratively recover the nonlinear relationship between covariates and survival time. Theoretically, the validity of measurement error correction and the estimation procedure are justified. Numerical studies show that the proposed method improves the performance of estimation and is able to capture informative covariates. The methodology is applied to analyze breast Cancer data provided by the Netherlands Cancer Institute for research.
