|
Abstract:
|
Longitudinal biomarkers and right-censored data are common in biomedical studies. Function cox model has been proposed to deal with functional covariates and right-censored data. However, it is not able to deal with sparsely and intermittently measured longitudinal biomarkers with measurement errors. The aim of this paper is to extend the framework of functional cox model to deal with longitudinal biomarkers. Specifically, we use functional principal component analysis to deal with the longitudinal biomarker, the results of which is then given as input to the cox model. The finite sample performance is illustrated by simulation studies and a real data application. The application to the Idiopathic Pulmonary Fibrosis data produces biologically meaningful results, which cannot be obtained using existing methods, suggesting the effectiveness of our proposed method.
|
