The Research Of Cure Model With Interval Censoring

A cure model is a useful approach for analysing failure time data in which some subjects could eventually experience and others never experience the event of interest.All subjects in the test belong to one of the two groups:the non-cured group and the cured group."case 1" interval censored data arise in many practical situations in which the event time of interest cannot be observed exactly,but is only known whether the failure event has occurred before or after a censoring time.Traditional statistical methods and models for censored-data are almost based on the hypothesis is always invalid.Dependent censoring has recently attracted a great deal of attention.This paper propose a cure model with independent and dependent interval censoring.First,we propose a cured model for“case 1”interval censored data.An EM algorithm is used to maximize the log-likelihood of regression parameter and survival function of the event time.Simulation results show that the proposed method can get efficient estimations.Then,here we propose a cure model for dependent interval censored data.We model the failure time's association with the observation time through a known copula function.An EM algorithm is proposed for nonpa-rametric estimation of the failure time distribution.Sensitivity analysis shows that the proposed method can achieve robust estimations.