Advancements in electrocardiogram signal classification algorithms

Abstract Investigating techniques for electrocardiogram (ECG) signal classification is essential in medicine, offering significant potential for the early detection and continuous assessment of cardiac disorders. This review article discusses the background context and current state of cardiovascular disease prevention and therapy, and also developments in computer-assisted ECG signal classification methods. It provides an overview of advancements in ECG signal categorization, including conventional machine learning methods, deep learning frameworks, hybrid models, and specialized methods such as flexible supervision strategies, spiking and hardware-aware neural architectures, and clinical knowledge-driven approaches. In addition, the performance and distinctive features of these models are outlined, emphasizing emerging research avenues and providing substantial technical assistance for future advancements in this field.