Chronic heart failure patients have an increased risk for frequent hospitalizations and mortality if adequate therapy is initiated too late. Clinical decision-making is based on data from different sources that include patients’ clinical data, ECG signals, and images from different modalities. For diagnosis and to define the optimal treatment strategy, the cardiologist has to combine this complex information. To automate the process, we develop AI-based decision-making using all available data to stratify the risk of hospitalization in patients with heart failure in the outpatient cardiology clinic. Timely identification of patients at high risk will allow for early intervention in the group and thereby may prevent hospitalization, reduce the consumption of hospital resources and improve patients’ quality of life. The project is part of the UvA Research Priority Agenda AI for Health Decision-Making, which is a collaboration between Faculties of Medicine, Science, Humanities and Law.