Accelerated Failure Model With Informative Interval Censored Data

People may often encounter interval-censored data in biostatistics?biomedicine?sociology and other research.At present,there have been many documents studied the problem of regression analysis of interval-censored data.However,most of them focus on the independent or non-informative censoring that the censoring time and the event failure time are independent,while the research on the dependent case is relatively few.In this paper,we will discuss the regression analysis of censored data in dependent intervals under the accelerated failure time model.We use Copula function method to characterize the relationship between interest failure time and observation time,and then obtain the likelihood function of observation data.Furthermore,we use spline function to approximate the nonparametric term of the likelihood function,and obtain the maximum likelihood estimation of regression coefficient.A large number of simulation studies show that the method we proposed has great performance under limited samples.We also apply this method to a set of practical data.