[Approximate entropy of electrocardiogram signals in atrial fibrillation].

Atrial fibrillation (Af) is a common arrhythmia and disorder of the rhythm of electrical activation of the cardiac atria. There is a high risk of thromboembolism in patients with Af. Therefore, it is an indication for anticoagulant therapy. The approximate entropy (ApEn) of electrocardiogram (ECG) signals was calculated in 29 subjects without arrhythmia and 12 patients with chronic Af. The mean ApEn in the group without arrhythmia was 0.347 +/- 0.119, and that of the chronic Af group was 0.556 +/- 0.169. ApEn was significantly higher in the chronic Af group than in the no arrhythmia group (p < 0.002). The average ApEn was significantly higher in the group not administered anti-arrhythmic drugs (0.687 +/- 0.033) than that in the group administered anti-arrhythmic drugs (0.516 +/- 0.209) (p < 0.05). We found that ApEn was effective in evaluating the nonlinear property and ApEn of ECG signals, enabling assessment of the randomness of chronic Af.