Acquisition Analysis And System Research Of Dynamic ECG Signal

As the number one killer in all disease, cardiovascular disease has been a serious threat to human life, the diagnosis and treatment of cardiovascular disease has become the most important issue on the medicine. However, ECG does not detect a transient or paroxysmal arrhythmias and myocardial ischemia, on the contrary, dynamic ECG can continuously record 24-hours ECG based on daily activities, which can compense the insufficient of ECG detection on transient or paroxysmal arrhythmias and myocardial ischemia, it is important for the prevention and diagnosis of heart disease.However,24-hours Holter monitoring systems need to manually record their daily activities, which caused not only the inconvenience for the recording subjects, but also the non-analytical of ECG data when they forget to record or lose the recording. This thesis obtain the acceleration signal through acceleration sensor to automatically detect the daily activities, which not only reduces the troublesome operation for manually recording the daily activities of subjects, but also has an important role in the treatment of rehabilitation exercise though controlling physical activity by the acceleration signal.Currently, remote monitoring Holter system has become a hot research, in which the user will not limit by the activities, the recording time and location. This thesis will use the ZigBee wireless comunivation technology because of its characteristic of low power, low cost and low whims, which is very suitable for the medical industry.Therefore, this thesis designed a wireless monitoring system based on daily activities. First, the system simultaneously collected the ECG signals and the acceleration signals, then transmissed the signals by ZigBee wireless technology, and finally display the signals by real-time on a computer. The accqusition of acceleration signals realized distinction and judgments of the daily behavior action, this paper used 4 kinds of behavior action, that still-standing, standing up, sitting down and bending, to distinct and judge.At present, researches on the static ECG has been higher results of systematic recognition, but the study of dynamic ECG analysis is still found no conclusion of the uniform system, the accuracy rate of detection and analysis of dynamic ECG is still less than on the static ECG analysis results. Therefore, there are still great prospects for development and the practical significance to realize the detection and analysis of dynamic ECG.On the dynamnic ECG anlysis, this thesis extracted the dynamic ECG feature under different methods, different leads and different noise interference. For the Ⅱ lead, we used wavelet transform and mathematical morphology method to dectect the R-wave, for the V3 lead, we dectected the P-QRS-T wave, and for the Ⅰ lead, we added the transform post-processing algorithms to improve the detection accurary, and contrast its detection result on motion artifact and EMG interference.Finally, this thesis verifies the relative accuracy of R-wave detection on the Ⅰ lead, and assessed the dynamic ECG signal on two signals of different relative accuracy and one signal of four kinds of ECG under daily activities through three kinds of parameters.