An Automatic Approach for Estimation of CPR Signal using Thoracic Impedance

Quality of chest compressions is an important measure with regards to cardiopulmonary resuscitation (CPR). For chest compression quality assessment, it is important to detect or estimate the chest compressions in heart rhythms. In this study, we proposed a new approach for estimation of the CPR during out-of-hospital cardiac arrest, and assessment of chest compression quality via CPR parameters. The CPR signal is estimated by combining candidate modes derived from an ensemble empirical mode decomposition (EEMD) based on frequencies of modes and thoracic impedance signal. The CPR parameters including compression numbers, compression rates, flow-time, and no-flow-time, from the estimated CPR are also evaluated with parameters derived from the reference CPR. Experiments show good agreements with high correlation between parameters by estimated and reference CRP signals, that demonstrate the performances of the proposed approach.