T-wave Alternans Detection Technique Based On ARM

Cardio-system disease is the most dangerous killer in the world. As a nonstationary phenomenon, T-wave alternans (TWA) is a variation in the amplitude or morphology of the T wave of the beat-to-beat ECG. Considerable scientific data support the potential value of T wave alternans(TWA) is ralated to ventricular arrhythmias and sudden cardiac death and provides a noninvasive and clinically useful marker for the risk of them.As the fast development of the digital signal processing and computer technology,the MTWA has been detected with high precision.Studying the TWA detection technique as the center,the paper presents the hardware and software based on ARM structure in the purpose of realtime ward.The ECG signal needs pretreatment before detecting the TWA,such as denoising and extracting characteristic parameters.Wavelet transform is chosen as the signal processing technology for its multi-resolution and the capability of describing the local characteristic of signal in both time and frequency domain.The raw ECG signal is decomposed into the different bands with the wavelet transform.According to the frequencies of ECG and the three major noise,we will able to achieve the elimination of the noise using the wavelet threshold denoising with the adaptive threshold value and eclectic threshold function;and then, we use Mexican-hat wavelet transform to detect R-peak of ECG signal based on its corresponding with the extreme of Mexican-hat wavelet transform;in the meantime,the position will be corrected through the extra tragedy,and the T matrix is given.And then the paper introduces the mechanism and the research progress of the T-wave alternans from clinical application and detection technique,and the spectral method,correlation method and modified moving average method are adopted to detect TWA by some ECG data from time and frequency domain.The spectral method has the strong capability of anti-noise,but it cannot detect the nonstationary TWA signal due to the need of the stationary ECG signal in the spectral technique.The correlation method has the stringent data requirements because of the influence of the respiration and other noises.Therefore we detect the signal by the specatral technique firstly,then,if the significant TWA is detected,we adopt the correlation method to confirm the alternans amplitude and time.At last,the influencing factors are analyzed to reduce the detection error.The paper designes the key part of the hardware and the software of TWA theoretical sample. The hardware includes the collecting module and the data processing module(external storage and communication interface circuits) in the core of the RISC ARM processor S3C44B0X.Based on the weak and the noisy signal,the paper designes the grading amplifier with the characters of high common mode rejection ratio and high input impedance to extract the effective signal.In addition,A/D circuit ensures the precise quantization,and the model machine communicates with the external environment through USB interface chip and UART in ARM.The software firstly introduces development environment and then presents the realization and flow chart of ECG analysis and processing procedures to monitor the whole system.