This invited talk, presented as part of a Biosignals subject at Universidad de Comillas, provides an overview of my research on electrocardiogram (ECG)-based biometric identification and its adaptation for atrial fibrillation (AFib) detection. The presentation explores the transformation of ECG signals into visual heatmaps (EKMs or ECMs) for user authentication and cardiac health monitoring, leveraging deep learning techniques. I will discuss how our methodology, initially designed for patient identification, was extended to AFib detection using convolutional neural networks (CNNs). The talk aims to illustrate the robustness of ECG-based identification and highlight the potential of deep learning in personalized healthcare and security applications.