An adaptive filter for the reduction of cardiogenic oscillations on esophageal pressure signals

In pulmonary monitoring, the processing and interpretation of esophageal pressure (Pes) signals is often complicated by strong cardiogenic oscillations (CGO). Here, the authors describe an adaptive filter that reduces CGO based on simultaneous recordings of the electrocardiogram. The algorithm is tested in a computer simulation and on patient data. In simulation, a reduction of the mean squared error introduced by the CGO by 87%, is achieved. In patient data, the filter reduces the power in Pes in a ±0.25 Hz band around the heart rate by 76%