A Method to Evaluate the Quality of Chest Compressions During Manual Cardiopulmonary Resuscitation

Chest compressions are an important treatment in cardiopulmonary resuscitation. In the process of using manual chest compression auxiliary device, the calculation of chest compression depth and frequency mainly depends on the change of accelerometer. However, using accelerometers to calculate the frequency of chest compressions is not reliable. In order to improve the accuracy of artificial chest compressions, this study proposed a method to evaluate the quality of manual chest compressions based on peak acceleration time during artificial chest compressions. The depth, frequency and acceleration changes during chest compressions were recorded and analyzed using manual chest compressions data acquisition system on human body model. The calculated peak acceleration time difference was basically the same as the actual peak acceleration time difference by using a Douglas-Puck algorithm to process the depth data. The random forest method was used to classify chest compressions with four different parameters. An accurate classification could be achieved when the compression depth, compression time and peak acceleration time were used as data sets. Using peak acceleration time difference instead of single chest compression time to classify chest compression modes with four different situations, nearly 97% accuracy was achieved. Therefore, this study suggested that the use of time difference of peak accelerations, which varies significantly, has greater advantages in the actual use of manual chest compressions in frequency judgment.