In order to predict cerebrovascular event occurrences, this work introduces a novel method that combines computational fluid dynamics (CFD), structural analysis, and machine learning (ML). The study presents a multidisciplinary approach to evaluate the risk of carotid atherosclerosis and cerebrovascular event prediction by utilizing both imaging and non-imaging data. The study uses blood-flow simulations and 3D reconstruction techniques to identify important properties of plaque that may indicate cerebrovascular events. The analysis shows high accuracy of the model in predicting these events and is validated on a dataset of 134 asymptomatic individuals with carotid artery disease. The goal of this work is to improve clinical decision-making by providing a tool that blends machine learning algorithms, structural analysis, and CFD. The dataset imbalance was treated with two approaches in order to select the optimal one for the training of the Gradient Boosting Tree (GBT) classifier. The best GBT model yielded a balanced accuracy of 88%, having a ROC AUC of 0.92, a sensitivity of 0.88, and a specificity of 0.91.