Ranking by marginal utility provides an efficient way to reduce the data from ultra-high dimension to portable size. In order to handle the complex big data in great variability, the statistic that can measure the nonlinear relationship between response and marginal predictor were extensively discussed recently. Comparing to ordinary regression analysis, it is more challenging when the response is the survival time with possible censoring in biological discovery or precision medicine. In this talk, we first introduce a dependence notation called Survival Ball covariance which can measure the dependency between survival time and covariates. We show that Survival Ball covariance is zero if and only if survival time and covariates are independent. We further propose a rank-based statistic, which is consistent to Survival Ball covariance. Using this statistic as the marginal utility, we present a strong independence screening procedure with the property of strong screening consistency, that is, the screening set converges to the active set with probability one. Its performance in finite sample size is evaluated via simulations and illustrated by the analysis of one real data.