| The exact automatic analysis for electrocardiogram (ECG) signal is the key to the diagnosis in the cardiac disease, the ECG analysis software can provide the clear ECG to doctor, enhance the accuracy of analysis and diagnoses are significant in clinic.In this paper, the methods of computer automatic analysis technology are studied. Combining with the development of Holter ECG monitor instrument, the functions of display, pretreatment, waveform identification, parametric computation and analysis of ECG are realized. The ECG automatic analysis software system which is composed of ECG display, pretreatment and analysis modules is designed successfully. This system possesses clinical application value.In the display module, both Holter ECG and MIT-BIH Arrhythmia Database signals can be read and displayed in the main user interface. Then we analyze the interferences of ECG, and adopt the method of single-scale and multi-scale wavelet transform to restrain the interferences rejection effectively, such as 50Hz power line interference and baseline wander. In the analysis module, the method based on wavelet transform on time-frequency analyzing is presented to identify QRS complex in this paper, and the P wave and T wave of ECG signal are detected based on the detection of QRS. The real time heart rate calculation and analysis of Heart Rate Variability (HRV), bigeminy and trigeminy are implemented by picking up the parameters of ECG signal, such as R-R interval and P-R interval et..The Holter ECG and MIT-BIH arrhythmia database are used to evaluate this analysis algorithm in this paper, and the measurement accuracy is high. To meet the needs of real time and dynamic, the ECG signal with no pretreatment is processed with this algorithm. The experiment result shows the detection precision is all above 99.8%.The ECG system can realize the functions of display, pretreatment and analysis in this paper. Especially, the detection algorithm is precise and stable. The system also can meet the needs of real time and has preferable potential application. |