Adaptive stimulus artifact cancellation in gastric myoelectrical signals

Previous studies have shown that electrical stimulation of the stomach with appropriate parameters is a promising approach for treatment of gastroparetic patients. The recording of gastric myoelectrical activity by serosal electrodes is often used to evaluate the effect of stimulation. However, the major problem with the measurements of gastric myoelectrical activity during gastric pacing is the stimulus artifacts which are often superimposed on the recorded signal and make spectral analysis unreliable. The transform domain adaptive filter has been used to reduce the stimulus artifacts but only with limited success. This paper presents an adaptive stimulus artifacts reduction method based on fast convergence frequency domain adaptive filtering, without requiring recorded “stimulus-only” reference signal. The results from both simulations and real signals show that the frequency domain adaptive filtering is able to cancel stimulus artifacts with fast convergence speed