Design And Implementation Of New ECG Detection Software System Based On All Phase Spectral Analysis And Filtering

With the increase of patients in cardiovascular disease, coupled with the level ofawareness of the increasing demand for health care, the traditional Holter detectiontechnology has been unable to meet the requirements of the people. In recent years,the ECG detection technology is developing and improving constantly, but lookingfor more efficient and accurate ECG detection technology has been the researchers’pursuit. ECG detection algorithm mainly includes two core parts which are ECGwaveform detection and de-noising, and the technology is mainly reflected in ECGworkstation system.Because ECG is always mixed with the baseline drift and frequency interferencesignals, which seriously affected the ECG waveform detection. Therefore, the primarygoal of ECG detection technology is de-noising to obtain the pure ECG. To addressthis problem, this paper presented an efficient method to remove the baseline drift andfrequency interference with FIR digital filter, which combination a high pass filterwith cut-off frequency of0.67Hz and a notch filter noising filter with frequency of50Hz and its harmonics to achieve the purpose. Experiments show that, the high passfilter with frequency of0.67Hz can completely filter the baseline drift of signal, andSaved multiplier of94.35%, while the notch filter is able to remove the frequency of50Hz and its harmonics in interference. Meanwhile, the combined filter has a goodanti-noise performance, low computational complexity, and it is also very efficientand feasible.For the phenomenon of that frequency interference of the actual ECG signal isnot always just at the frequency of50Hz, this paper presents a novel method based onall phase filter design, the filter is combined with a0.05Hz~250Hz band-pass filterand a all phase notch filter in any point. The key technology is to estimate the exactfrequency of interference, and design a all-phase specific frequency notch filter.Experiments show that the combinated filter can not only remove the baseline drift,but also can notch interference signal according to its precise frequency in the actualcircumstances. Therefore, the method will have broader applicability and higherpractical value. After a series of pre-processing in the ECG signal, the paper proposed aalgorithm for detecting QRS waveform. Experimental results show that the algorithmcan accurately detect the location of P wave, QRS wave group and T wave, falsedetection and missed situation do no exist.Finally, ECG workstation system achieved ECG detection algorithm of the paper.The system includes five sections which are12-lead ECG, ventricular late potentials,vector cardiogram, heart rate variability analysis and high-frequency ECG. The papermade a detailed introduction which focused on the theory, software design andimplementation of the first four parts.