Detection of Atrial Fibrillation via Analysis of Mean, Entropy, and Heart Rate Parameters Using Electrocardiogram Signals

Atrial Fibrillation (AF), commonly referred to as AFib in clinical practice, is the most prevalent form of sustained cardiac arrhythmia and represents an important challenge in cardiovascular diagnosis and long-term patient monitoring. AF occurs when the normal coordinated electrical activity of the atria is disrupted, resulting in disorganized and irregular propagation of excitation waves and, consequently, an abnormal ventricular response. Accurate characterization of electrocardiographic alterations associated with AF is therefore essential for improving automated detection and supporting reliable clinical assessment. In this study, electrocardiogram (ECG) signals obtained from adult subjects are comparatively analyzed to distinguish healthy cardiac activity from patterns associated with Atrial Fibrillation. The dataset employed in the investigation was compiled from records provided by the American and European Heart Associations and includes subjects from different age groups. Demographic analysis indicates that fewer than ۵% of the participants belong to the ۴۰–۵۰-year age group, whereas approximately ۵–۱۵% are around ۸۰ years of age, reflecting the increased prevalence of AF among older individuals. A quantitative analysis of heart-rate behavior reveals that patients with Atrial Fibrillation exhibit substantially more irregular temporal dynamics and generally higher heart-rate values than healthy subjects. In AF cases, heart rates are frequently concentrated within the range of ۸۰–۱۰۰ beats per minute (bpm), with a considerable number of episodes exceeding ۱۰۰ bpm. In addition to heart-rate analysis, statistical and signal-based features are extracted from processed ECG segments to characterize differences between the two groups. The results demonstrate that AF signals exhibit a lower mathematical mean while presenting considerably higher entropy and energy values than ECG signals obtained from healthy individuals. These findings indicate that statistical characteristics such as mean, entropy, and signal energy can effectively capture the increased irregularity and complexity associated with Atrial Fibrillation. Consequently, the identified features may provide useful discriminatory information for the development of automated ECG-based AF detection and classification systems, particularly in continuous cardiac monitoring and computer-assisted diagnostic applications.

کلیدواژه ها:

Atrial Fibrillation ، mean ، Entropy ، Energy ، Heart Rate ، Electrocardiogram (ECG)