Proceedings of The Physiological Society

Physiology 2012 (Edinburgh) (2012) Proc Physiol Soc 27, PC183

Poster Communications

Human skin microcirculation assessment by wavelet transforms and detrended fluctuation analysis

H. Ferreira2, C. Fernandes2, P. Pinto1, L. A. Monteiro Rodrigues1,3

1. CBIOS UDE, U Lusofona FCTS, Lisboa, Portugal. 2. IBEB, U Lisboa FC, Lisboa, Portugal. 3. Dep Pharm Sciences, U Lisboa FF, Lisboa, Portugal.


Wavelet analysis from human skin's microcirculation signals obtained by Laser Doppler Flowmetry (LDF) shows components at characteristic frequency ranges related to the heart (0.6-2Hz), to respiration (0.15-0.6Hz), to the vessel wall myogenic activity (0.052-0.15Hz), to sympathetic activity (0.021-0.052Hz) and to endothelial metabolic activity (0.0095-0.021Hz). These wavelets components additionally show amplitude modulation in time and frequency, and may also hold important information regarding blood flow physiology. Detrended Fluctuation Analysis (DFA) is an interesting method to study blood flow fractal properties. Analyses of wavelet component amplitudes and DFA-derived alpha exponents have typically been used separately. Here we applied a combined approach of these two methods in order to get a cleared picture of microvascular blood-flow regulation. The study involved 9 female healthy young subjects (age=20.3+4.0 years) giving previous informed written consent. All procedures fully respected Helsinky principles and respective amendments. LDF measurements were recorded for 30 minutes - 10 min baseline; 10 min after a perfusion restriction of the ankle with a cuff, and 10 min recovery. Data sampling was 32Hz and, after data segmentation in the 3 time segments, analysis involved a home-built MATLAB script based on MATLAB's wavelet toolbox and DFA algorithm. Frequency components from wavelet decomposition were analyzed regarding amplitude ratios (mean amplitude of each component over the total signal amplitude) and alpha exponents. Comparison between data segments (baseline, perfusion restriction, and recovery) was done regarding amplitude ratios and alpha exponents for each frequency component by using paired t-test or Wilcoxon test, accordingly. Data suggest that each frequency component has a distinct amplitude ratio, with the metabolic and sympathetic components showing the highest values. Additionally, these components show the highest DFA alpha exponents, which translate a non-stationary correlated signal of the fractional Brownian motion type. During perfusion restriction, a significant increase of the amplitude ratios and alpha exponents of the heart, respiratory and myogenic components, which are expected to follow vasodilation, happens. Additionally, a decrease in metabolic activity is observed (NO release from the endothelium?). After cuff release, recovery of heart and respiratory components is observed, particularly in the alpha exponents. No significant changes were observed for the sympathetic component, probably due to the reduced sample size. The combined approach of wavelet analysis and DFA may provide a powerful way to analyze vasomotion in LDF signals. In particular, the method could differentiate both myogenic and endothelium responses in microcirculation territories.

Where applicable, experiments conform with Society ethical requirements