Feature extraction of myocardial infarction based on lead I ECG / 生物医学工程学杂志

ABSTRACT

This paper presents a myocardial infarction feature extraction algorithm, and whereby using single-cycle signal of electrocardiogram (ECG) lead I, we solved the problem of inconvenience in using multi-lead ECG. Firstly, the noise is eliminated from the signal of ECG lead I, and then, wavelet packet algorithm is introduced to extract the main frequency band of QRS complex and T wave, and reconstruct the QRS complex and T wave shape so as to determine the original and end positions of ST segment. Finally, ST segment is decomposed and the waveform characteristics of myocardial infarction from ECG lead I signal is obtained based on the analysis of wavelet's multi-resolution. Simulation results showed that the algorithm achieved a high recognition rate, which provides fundamental principle for the detection of myocardial infarction by using the ECG lead I signal.