Continuous monitoring of patients suffering from cardiovascular diseases and, in particular, myocardial infarction (MI) places a considerable burden on health-care systems and government budgets. The rise of wearable devices alleviates this burden, allowing for long-term patient monitoring in ambulatory settings. One of the major challenges in this area is to design ultra-low energy wearable devices for long-term monitoring of patients’ vital signs. In this work, we present a real-time event-driven classification technique, based on support vector machines (SVM) and statistical outlier detection. The main goal of this technique is to maintain a high classification accuracy while reducing the complexity of the classification algorithm. This technique leads to a reduction in energy consumption and thus battery lifetime extension. We validate our approach on a well-established and complete myocardial infarction (MI) database (Physiobank, PTB Diagnostic ECG database [1]). Our experimental evaluation demonstrates that our real-time classification scheme outperforms the existing approaches in terms of energy consumption and battery lifetime by a factor of 3, while maintaining the classification accuracy at a medically-acceptable level of 90%.