Analytical Study On ECG Signal Based On Wavelet Transform And Hardware Implementation

The heart is one of the body’s vital organs,the related diseases caused by the heart have become the main diseases endangering human health.As a noninvasive and effective diagnostic method,electrocardiogram,which has an important reference value in clinical diagnosis.The ECG signal contains a lot of information that characterizes the physiological status of the heart,ECG signal analysis and automatic diagnosis technology plays a key role in disease diagnosis,which is helpful to the application and development of portable ECG monitorirng equipment and promotes telemedicine services.This paper mainly focuses on the ECG signal preprocessing and characteristic waveform detection.Based on the wavelet transform algorithm,ECG signal preprocessing algorithm and characteristic waveform detection algorithm are designed.The hardware architecture of ECG signal preprocessing and characteristic waveform detection algorithm is proposed and implemented on FPGA device.Firstly,this article studies the wavelet transform algorithm.According to the noise characteristics of ECG signal,a denoising algorithm based on lifting wavelet threshold and moving average filtering is proposed.In order to remove the power frequency noise and EMG noise,the improved compromise threshold function is used to filter the noise of the wavelet coefficients.In order to remove the baseline drift,the moving average filter algorithm is used to filter the approximate coefficients of the highest scale to realize the removal of noise.The next step is to detect the characteristic waveform after signal preprocessing.In the aspect of characteristic waveform detection,this paper studies the detection principle of ECG signal mutation point.The characteristic waveform detection algorithm based on wavelet transform is improved to realize the detection of QRS complex.The R waveform position is determined by the modulus maxima of wavelet coefficients.Through the dynamic threshold and auxiliary strategy of characteristic waveform detection,the algorithm effectively improves the detection rate of R wave.Secondly,the hardware architecture is used to design the ECG signal preprocessing and characteristic waveform detection,which is implemented on FPGA hardware,and the implementation scheme of each key module is described in detail.In order to avoid the problem of boundary distortion and time delay in the data processing,the pipelined processing structure is adopted to implement the wavelet transform,and it improves the computational efficiency.In the preprocessing module,the circuit design includes lifting wavelet transform based on Sym4 wavelet,median estimation and moving average filter.In the characteristic waveform detection module,the circuit design mainly includes biorthogonal quadratic B-spline wavelet transform,modulus maxima detection,zero-crossing detection,and the above modules were described in RTL and conducted functional simulation.Finally,this paper validates the design based on Altera FPGA hardware,further ensures the correctness of ECG signal preprocessing and characteristic waveform detection circuit and the feasibility of RTL code.The design of this paper is evaluated through MIT-BIH ECG database,the result shows that the system has high R waveform detection rate and practical value.