Denoising of Diaphragmatic Electromyogram Signals for Respiratory Control and Diagnostic Purposes

Diaphragmatic electromyogram (EMGdi) signals give important information about the respiratory muscle pump, can be used as an indicator of neural respiratory drive, and have been postulated as a method of designing neurally-activated intelligent ventilators. However diaphragmatic EMG signals measured with an esophageal catheter tend to be contaminated by electrical signals from the heart-electrocardiogram (ECG). This paper presents a novel method of rapidly separating and enhancing the Electromyogram signals from the combined EMG and ECG signals recorded from an esophageal catheter based sensor. Independent Component Analysis (ICA) is used to separate the EMG and ECG signals, then further processing is used to extract the frequency of the patient's breathing and the relative magnitudes of diaphragmatic muscle activity. These signals have two applications, firstly in artificial ventilator systems and as a diagnostic tool for health professionals.