Modeling the 'digital twin' of a patient's heart has gained traction in the last years and helps to understand the pathogenic mechanisms of cardiovascular disease to pave the way for personalized therapies. Although a 3D patient-specific model (PSM) can be obtained from computed tomography (CT) or magnetic resonance imaging (MRI), the fiber orientation of cardiac muscle, which significantly affects the electrophysiological and mechanical characteristics of the heart, can hardly be obtained in vivo. Several approaches have been suggested to solve this problem. However, most of them require a considerable amount of human interaction, which is both time-consuming and a potential source of error. In this work, a highly automated pipeline based on a Laplace- Dirichlet-rule-based method (LDRBM) for annotating fibers and anatomical regions in both atria is introduced. The calculated fiber arrangement was regionally compared with anatomical observations from literature and faithfully reproduced clinical and experimental data.