Development of a minimally invasive microvascular ischemia monitor: noise and artifact extraction

Long-term postoperative pH monitoring of transplanted tissue for ischemia produces signals contaminated with noise and artifacts of various origins. Clinical interpretation of these pH recordings is difficult. The objective of our research is to determine the optimal filtering parameters for tissue pH signals to extract and reject the artifacts. The noise components present in long-term pH recordings were identified, extracted, and analyzed. Frequency ranges of identified artifact types were determined. Parameters were defined to effectively characterize each of the components by frequency, amplitude, and peak to peak slope. Based on the artifact analysis, it was concluded that to achieve clinical usefulness, only motion and position artifacts need to be rejected from the signal, while the influences of the “24 hour component” and drift are insignificant