analysis of spatial survival data using copula functions

Many survival data analyses aim to assess the effect of different risk factors on survival time. when these survival times are correlated, using the copula functions is common to incorporate the data correlations. in some studies, the correlation between survival times is related to their spatial locations. one may use the copula functions to model the spatial dependence of the data. spatial data are usually positively correlated, and the correlation decreases as the spatial distance between units increases. so, a copula function can be suitable if it covers positive correlations between variables. in addition, the copula results in independence and maximum correlation when the spatial distance between units goes to infinity and zero, respectively. in this talk, we present a useful spatial copula function, which has a proper fit to survival data because of its asymmetric property. since the copula approach models the marginal density functions and the dependence structure separately, a two-stage method is used to estimate the regression and dependence parameters separately. in a simulation study, we investigate the performance of the proposed model‎. ‎next‎, ‎this model is applied to analyze a real dataset on acute myeloid leukemia. finally‎, ‎a brief discussion is presented.