Research And Application Of Key Technology In Intracardiac ECG Analysis

The collecting of internal ECG signal is a kind of invasive examination. In internal ECG signal, P wave is the most prominent wave of all ECG waves instead of QRS complex. P wave reflects the potential distribution of ventricular depolarization. It is the perfect time to measure blood impedance right before P wave (atria contraction), i.e. the end of diastole. At that moment the right atrium is completely filled with blood and the impedance signal is least influenced by the proximity of the atria wall. Furthermore, at this moment low flow exists in the right atrium and hence blood viscosity values will be the highest. There are clear evidences that blood impedance can be correlated with whole blood viscosity. While by measuring blood viscosity it is possible to detect the level of inflammation in blood and the pre-thrombotic risk; furthermore blood viscosity determines also part of the hemodynamic load for the heart. Therefore, accurately detecting the positions of P waves and finding the corresponding blood impedance has great significance for the diagnosis and therapy of cardiovascular disease. Heart rate variability can reflect the activity of the autonomic nervous system and quantitatively evaluate the tension and balance between cardiac sympathetic nerve and vagus nerve, which has been a valuable index for prediction and diagnosis of cardiovascular disease. It has been of great value for both clinical application and physiological research, which makes HRV a hot topic in the field of cardiovascular disease research.Firstly, the paper summarizes the current situation of the development in ECG automatic analysis and compares the characteristics of the existed ECG signal detection algorithms. Finally, the typical time-domain method is chosen by tradeoff between the processing speed and the detection accuracy. The method firstly emphasizes the P waves by a series of simple algorithms, and then detects the P waves by using an adjusted threshold on the enhanced signal, finally adding the recheck process which assures the sensitivity and positive predictivity to some extent. After detecting the P waves, the algorithm also detects the onset and offset of P waves. This paper also introduces the common used methods of HRV analysis in detail, and realizes most of the methods in time domain and frequency domain.The algorithms used in the paper achieve good performance by testing on the internal ECG signals supplied by Martil Instruments. The sensitivity and positive predictivity are both above 98%, and the HRV results are compatible with those of Kubios software from internet.