The Applications Research Of The Fractional Wavelet Transform In The ECG Processing

Biomedical signals,belongs to the low frequency weak signal in strong noise background,it is composed of complex organissms occurred unstable natural signal by itself,the charactertics,the detection and the processing technology,are different from general signal.All kinds of signal needs to be extracted and processed, such as EEG signals, EOG, gastric electrical and other physiological signals, are an important diagnosis basis of the patient’s detection and treatment.ECG signals and other biomedical signals, all belong to low frequency weak signal, when testing,it is often susceptible to be interfered by the surroundings. Even the same body of biological signal detection also can have different results under different circumstances.So the ECG signals is characterized by a weak signal,strong noise, strong randomness and so on.ECG signal is a synthetic reflection of the heart electricity on body surface. In clinical practice,it has important significance to the diagnosis of heart disease.In real life, people use ECG to record the heart produced by bioelectric current status,clinical doctors can use the ECG to assess the patient’s heart condition,and further diagnosis.However, the power frequency that produced by instruments, lead wires and the human body, baseline drift,EMG interference,etc, are severely interfered with the ECG records.To filter out these interference and pick up the useful information are of great significance to signal detection and processing.ECG signal now is a non-stationary random signal. Because of intuition and ease of detection, it has caused more and more researchers, concern. This paper studies the properties of the wavelet transform, the fractional Fourier transform and the principle method for processing signals. For non-stationary signals, the wavelet transform and time-frequency analysis combined together, that is frequency Focus of fractional Fourier transform combined with multi-resolution analysis of wavelet transform, Compared to the traditional wavelet only using transform threshold method. In this paper use ECG domain in the best p-order fractional Fourier transform, then process the wavelet transform. This method is more suitable for the extraction and elimination of noise. It can broaden the scope of application of fractional Fourier transform and enrich theory of the fractional Fourier transform. So the discrete Fourier transform is applied Fractional wavelet transform domain filtering for signal detection and other areas are of great significance. Secondly, it can use the modulus great extremes of the wavelet transform to detect and correct the detected R-wave positive and negative R waves, and use the missed, wrong pick compensation strategies to improve the accuracy of the R wave. Q-wave and S-wave are adopted a priori knowledge of the Combination and windowing methods to test.After the algorithm is complete, with MIT-BIH database multiple pieces of data simulation, the results show that using fractional wavelet transform of ECG signal filtering highest signal-to-noise ratio can reach up 29.745db;Using the modulus great extremes method of QRS wave detection, accuracy can reach up 99.81%.The results show that the algorithm has high accuracy.The hardware design is based on embedded processsor S3c2410 core.Under the WinCE system,it designed the corresponding filter and testing program.By the test,the algorithm is feasible on the system.