Feature Point Detection On Signals From Epicardial Mapping

Cardiac mapping is a major technique of researching the mechanism of cardiac electrophysiology. This research benefits from epicardial mapping signals (in sinus rhythm and atrial fibrillation) acquired from canine models by 128-channel epicardial potential mapping system, which is designed by Fudan University. The signals are analyzed with Matlab software. After signal preprocessing, different algorithms including Amplitude threshold, Slope threshold and Botteron’s approach were applied to detect the feature points (activation time) of the epicardial signals, then the quantity of feature points was calculated, also including activation frequency and etc. Besides, this research compared these results with that derived from the dominate frequency method and statistically analyzed the difference. The final result shows that, considering both sinus rhythm and atrial fibrillation, Botteron’ approach combined with slope threshold can obtain the best effect, i.e. the best repeatability and accuracy. The conclusion based on this research can be used to illustrate and explore the atrial electrical activities in the form of figure and diagram.