Survey on Explainable AI Based Meta-Learning to Enhance Patient-Specific ECG Anomaly Detection

Cardiovascular disorders are becoming increasingly prevalent and represent a significant public health challenge. Today, computerized ECG diagnostic systems often fail to identify rare but important critical heart problems. This is mostly because ECG datasets are not always balanced. Electrocardiograms (ECG) are essential for evaluating cardiac health and detecting irregularities, and their early identification plays a pivotal role in improving patient outcomes. This study examines this field by developing advanced machine-learning algorithms and Explainable AI (XAI) techniques for automated ECG classification. By employing meta-learning approaches, we harness normal ECG data to train models that can effectively identify deviations indicative of rare and critical cardiac function. This method notably enhances Detection effectiveness, but also promotes transparency in decision-making, ultimately leading to better patient care and outcomes in cardiology.