2.5. Statistical analysis

Continuous variables with normal distribution are presented as means with standard deviations and were compared using the Independent-Samples T test, whereas variables with non-normal distribution are presented as medians with interquartile ranges and were compared using the Mann-Whitney U test. Categorical variables were compared using the Chi-square test. Receiver operating characteristic (ROC) curves were plotted to determine the optimal infarct size cutoff-points for predicting different types of pontine lacunar infarction. The “golden” criterion for lacunar infarction in the pons is defined as a lesion not extending to the anterior surface of the pons. We assessed discrimination by calculating the area under the ROC curve (AUC). An area of 1 implies a test with perfect sensitivity and specificity, while an area of 0.5 implies that the model’s predictions are no better than chance. Optimal cutoff-points for size were determined by maximizing the Youden index. Statistical analyses were performed using SPSS 20.0 software. P values of < 0.05 were considered statistically significant.