چكيده به لاتين
The processing of heart signals and separating their parameters is used for medical diagnosis. In this study, a method for the diagnosis of HRV using heart signal processing is provided to help the prescriber during the treatment and diagnosis of the disease. In order to detect the HRV using cardiac signals, the first is to extract the R_R intervals of the ECG signal for the formation of HRV signals. From each HRV extracted signal, the time, frequency and nonlinear characteristics are extracted. Using the identification of these features to identify the different types of arrhythmia that is used to detect coronary arteries. Design and implementation of a suitable neural network for detecting HRV from heart signals was presented and displaying outputs on diagrams and implementations using MATLAB neural network simulators. Then the comparison of the proposed method and the model with similar models was carried out in terms of low error estimation and prediction accuracy coefficient. In this research, the proposed model is a new model among the other models presented in terms of the low percentage of predictive errors and the high precision coefficient among traditional HRV diagnostic models from heart signals. From the obtained error coefficient, it can be said that among the suggested models of this method can be a good method.
Keywords: HRV Diagnosis, Neural Network, Heart Signals, Arrhythmia, Genetic Algorithm
