Variability Investigation of Blood Pressure Estimates Extracted from Photoplethysmogram

Unobtrusive methods to monitor blood pressure variability emerges a significant biomarker of cardiovascular control mechanisms. Short-term blood pressure fluctuations might have a prognostic relevance, predicting organ damage and cardiovascular events. There are evidences of positive correlation between increased blood pressure variability, higher average blood pressure level and age. However, until now increased blood pressure variability indexes are not utilized in the clinical practice. One of the preventing factors is unavailability of long-term unobtrusive methods for continuous blood pressure registration. Goldstandard 24h blood pressure monitoring with 15 min intervals provides information on blood pressure fluctuations occurring in the “very low” frequency range, whereas the higher-frequency components of blood pressure variability go undetected and can only be assessed with a beat-to-beat measurement technique. The pulse volume waves unobtrusively registered as Photoplethysmogram is an attractive alternative to represent ongoing changes in the body hemodynamics. The primary goal of the research was to investigate correlations between the variability of blood pressure parameters and indexes extracted from the model-based analysis of Photoplethysmogram signal. The algorithms for tracking the systolic edge slopes of Photoplethysmogram were investigated primarily. Parametric models (linear and hyperbolic tangent) were fitted continuously over time to detected systolic edges of the hemodynamic pulses. Algorithms for tracking the extracted parameters over time were developed in Matlab software. The preliminary results of experiments show small association between extracted Photoplethysmogram parameters and blood pressure variation during the tilt table test.