Improved signal processing techniques for the analysis of high resolution serosal slow wave activity in the stomach

High resolution electrical mapping of slow waves on the stomach serosa has improved our understanding of gastric electrical activity in normal and diseased states. In order to assess the signals acquired from high resolution mapping, a robust framework is required. Our framework is semi-automated and allows for rapid processing, analysis and interpretation of slow waves via qualitative and quantitative measures including isochronal activation time mapping, and velocity and amplitude mapping. Noise removal techniques were validated for raw recorded signals, where three filters were evaluated for baseline drift removal and three filters for removal of high frequency interference. For baseline drift removal, the Gaussian moving median filter was most effective, while for eliminating high frequency interference the Savitzky Golay filter was the most effective. Methods for assessing slow wave velocity and amplitude were investigated. To estimate slow wave velocity, a finite difference approach with interpolation and smoothing was used. To evaluate the slow wave amplitude and width, a peak and trough method based on Savitzky Golay derivative filters was used. Together, these methods constitute a significantly improved framework for analyzing gastric high resolution mapping data.