Ahmad Keshtkar
1*, Negisa Seyyedi
2, Shabnam Afkari
3, peyman Sheikhzadeh
4*, Seyyed Hossein Rasta
51 Associate Professor, Department of Medical Physics, School of Medicine, Tabriz University of Medical Sciences, Tabriz, Iran
2 Department of Medical Informatics, School of Advanced Technologies in Medicine, Tehran University of Medical Sciences, Tehran, Iran
3 Department of Telecommunications Engineering, School of Electrical and Computer Engineering, University of Tabriz, Tabriz, Iran
4 Department of Medical Physics, School of Medicine, Tabriz University of Medical Sciences, Tabriz, Iran
5 Assistant Professor, Department of Medical Physics, School of Medicine, Tabriz University of Medical Sciences, Tabriz, Iran
Abstract
Background: There are varieties of electrocardiogram-based methods to predict the risk of Ventricular tachycardia in patients. New extracted features from the signal averaged electrocardiogram and its wavelet coefficient as a distinction’s index are used in this study. Methods: Signals of orthogonal leads from 60 myocardial infarction patients (MI) with or without the history of ventricular tachycardia were selected from the national metrology institute of Germany (PTB diagnostic database). They were filtered and the discrete transformed wavelet was exerted on them. New and conventional features introduced in this study were extracted from signal averaged electrocardiogram and its wavelet decompositions. Results: Extracted features: QRS-d, Entropy-w, Maxhist and ZeroC has acceptable statistically criteria (p-value <0.05) for mentioned groups, comparing QRS duration ,in MI patients which is longer than MI + VT, and for other features it is Vice versa. In wavelet decomposition analysis, the entropy feature has higher precision for detection and diagnosing MI and MI+VT. Conclusions: Entropy of wavelet coefficients is a useful feature in distinguishing myocardial infarction patients with or without ventricular tachycardia.