| Popis: |
The Esophageal Pressure (Peso) signal can be used to monitor the respiratory mechanics of critically ill patients in the Intensive Care Unit (ICU), and has been successfully used in guiding mechanical ventilation strategies to improve patient outcomes. However, cardiogenic oscillations (CGOs) are a major source of interference, which not only makes it challenging in interpreting the patient’s respiratory mechanics, but can also cause false triggers in the mechanical ventilator resulting in a patient-ventilator asynchrony. In this thesis, we present a Peso enhancement scheme using Ensemble Empirical Mode Decomposition (EEMD) to suppress CGO interference. The proposed method was applied to synthetically generated Peso signals as well as real-world Peso signals from mechanically ventilated ICU patients. The proposed technique has been shown to significantly reduce the amplitude fluctuations caused by CGOs. The technique’s performance has been assessed through Face Validation by our collaborating clinicians, and is found to be suitable in not only suppressing CGO, but also extracting CGO from clinically acquired Peso signals. |
