Improving Stroke Risk Prediction in the General Population: A Comparative Assessment of Common Clinical Rules, a New Multimorbid Index, and Machine-Learning-Based Algorithms

Complex relationships of various comorbidities uncovered using a ML approach for diverse (and dynamic) multimorbidity changes have major consequences for stroke risk prediction. This approach may facilitate automated approaches for dynamic risk stratification in the significant presence of multimorbidity, helping in the decision-making process for risk assessment and integrated/holistic management.