Statistical analysis. Baseline characteristics ,use of cardiopulmonary bypass ,and the atherosclerosis score were compared in patients with and without post-CABG stroke. Cardiac embolic sources and post-operative hemodynamic impairments were compared in patients with post-CABG atherosclerotic and other stroke. Four atherosclerotic diseases were defined according to arteries that were considered when defining significant ( _50% steno occlusive ) atherosclerosis: ECCA disease ,extracranial cerebral atherosclerosis, intracranial cerebral atherosclerosis ,and extracranial and/or intracranial cerebral atherosclerosis. Associations between post-CABG stroke and each atherosclerotic disease were also compared .In univariate analysis ,we used the Pearson chis-quaretest with the Fisher exact test to examine categorical variables and the Student t test or the Mann- Whitney U test to explore continuous variables .Multiple logistic regression analysis was performed to estimate independent contributions of variables to the development of post-CABG stroke .Variables with p values _0.1 by univariate analyses were candidates for inclusion in multiple logistic models. A backwardel imination process was used to develop the final multivariable model, and adjusted odds ratios with95% confidence intervals (CIs) were calculated. A 2-tailed p value _0.05 was considered statistically significant. All statistic alanalyses were performed using SPSS software version 14.0 (SPSSI nc., Chicago ,Illinois).