The Research On ECG Signal Detection And Automated Analysis Techniques

Cardiovascular and cerebrovascular diseases have been always one of the major diseases that can threaten human health and life safety, while the ECG is a reflection of cardiac activity status in the body surface, and it is an important way to diagnose diseases. It is the premise to classify and identify the heart disease correctly by preprocessing and waveform detection and parameter extraction technique of the ECG signal. So, the research on the ECG automatic analysis technology becomes one of the hot spot in the field of signal processing.This paper designed an ECG front-end detection circuit, and it can carry on analyzing through collecting the real-time ECG signal by certain equipments. And it made a detailed presentation on the preprocessing and the waveform detection technology of the ECG signal. Especially for researching the algorithm of the ECG preprocessing technology, it obtained good results. The main works of this paper were presented as follows:1、Firstly, this paper started with a brief introduction on the generation mechanism and introduced characteristics and the sources of interference of the ECG signal. ECG signal belongs to weak signal in the low-frequency so that it is easy to interfere by all kinds of noise, such as power line interference, baseline wander, the EMG interference and so on. According to the characteristics of the different noise, this paper introduced the different methods to suppress the interference.2、In connection with characteristics of the ECG noise, this paper analyzed the detection requirements of the ECG signal. And it designed the front-end detection circuit of the ECG signal through designing the reasonable filter and amplifier parameters.3、It studied the ECG signal denoising algorithm based on wavelet transform and improved the wavelet threshold by using weighted average method based on Donoho fixed threshold forms, and it was carried out the simulation analysis of the algorithm by using the data in the MIT-BIH ECG database and real-time detection data. The experimental results showed that the algorithm could effectively remove the high-frequency interference and some baseline wander in the ECG signal.4、It analyzed the characteristics and physiological significance of the various ECG waveform and studied the detection algorithm. For the detection of the QRS complex, this paper introduced the differential threshold method and wavelet transform method. The paper used a combination detection method based on spline wavelet and differential method. By the data test in MIT-BIH ECG database, it has a higher accuracy of the R wave detection, the starting point of Q-wave and the ending point of S-wave are accurate positioning at the same time.5、It extracted two characteristic parameters of RR interval and width for QRS complex on the basis of the detected QRS complex, and calculated the heart rate of the ECG signal. Then according to the characteristics of the arrhythmia signal, this paper defined the pathological features and the automatic classification threshold based on the ten kinds of arrhythmias diseases, and designed Arrhythmia classification identification algorithm based on the branching logic.