چكيده به لاتين
Sudden changes in cardiac signals and followed by that, heart attack and cardiac arrest are one of the most common causes of sudden deaths. Due to the importance of this subject, various tools and methods have been devised to investigate the way of heart function in modern medicine, such as cardiac signal recording (Electrocardiogram (ECG)), angiography, and so on. Because the analysis of recorded images is the responsibility of expert and since that usually severe changes in cardiac function occurs suddenly and almost at the time of accident, the probability of error in diagnosis will be high and in most cases patients' lives will be at risk of death. For this reason, many studies have been carried out in this field in order to smartening and increase the detection accuracy. Sudden cardiac death (SCD) is one of the leading causes of death among people. A proposed method is presented for predicting SCD based on ECG signals by using wavelet packet transform (WPT), signal processing technique, homogeneity index (HI), nonlinear measurement for time series signals and neural network classification algorithm. The effectiveness and usefulness of the proposed method are measured and evaluated by using a database of ECG data. In the proposed method, HI values are calculated for 512 nodes from the first level of decomposition to eight levels of WPT decomposition for each 1 minute interval and subject. This analysis improves the effectiveness of proposed method and computational efficiency.
