In this article is reported a study on the optimal complexity of the long electrocardiogram (ECG) signal. At first, the original signal sequence was symbolized, then its complexity was calculated by Lempel-Ziv algorithm, and different factors of the complexity were discussed, and three different kinds of ECG signals, i.e., normalcy, angina and myocardial infarction, were studied. The results indicate that the optimal threshold-value to symbolize the original signal and the optimal signal length are two main factors that affect the complexity value, and the optimal complexity value can effectively distinguish normal ECG and patients' ECG.
Algorithms ; Angina Pectoris ; diagnosis ; Biomedical Engineering ; Diagnosis, Computer-Assisted ; methods ; Electrocardiography ; methods ; Humans ; Pattern Recognition, Automated ; methods ; Signal Processing, Computer-Assisted