Automatic Detection Algorithm for Chest Compressions Signal with Classification Algorithm
Transthoracic impedance (TTI) has been demonstrated to be a potential indicator to monitor the quality of chest compressions (CCs) during cardiopulmonary resuscitation (CPR). However, TTI signals are challenged by noise artifact from multiple sources, such as ventilations and baseline drift. Practically, it is very essential to accurately detect the peak-to-trough of the complicated TTI signals. However, nowadays, there is no method to solve the problem. In this paper, Extrima search with niche technology was used to search the peak-to-trough of TTI signal. We select 2 features to judge the potential peaks and troughs in order to remove the false ones. Besides, we designed a LDA classifier for recognizing the compression and ventilation waves. The experimental results show that this method in this paper can precisely recognize the real peaks and troughs of TTI signals which include some false ones.