ECG analysis involves the visual inspection of signal patterns, a practice that is inherently complex and reliant on the clinician's expertise. Despite the integration of computer algorithms to assist this process, such systems often lack the reliability and trust required by medical professionals due to their opaque mechanisms. Recent advancements in machine learning have yielded models with impressive diagnostic accuracies. However, these often fall short in terms of explainability and alignment with clinical practices. Vision Transformers (ViTs) offer a novel approach to interpreting electrocardiograms (ECGs) by leveraging advanced machine learning techniques to emulate the diagnostic processes used by clinicians. This paper explores the potential of ViTs to enhance the interpretability and accuracy of machine-assisted ECG diagnostics. Our solution is trained on the same visual arrangements of ECG leads utilized by clinicians, so our ViT could bridge the gap between automated analysis and clinical intuition. Employing an attention mechanism akin to human cognitive processes, ViTs promise a more transparent and reliable diagnostic tool. Building on previous research, this study aims to deepen our understanding of how artificial attention mechanisms can improve the accuracy and trustworthiness of ECG interpretation, thereby advancing the integration of machine learning in clinical diagnostics.