Evaluation of arterial stiffness by finger-toe pulse wave velocity: optimization of signal processing and clinical validation

摘要

Background:Carotid-femoral pulse wave velocity (PWV) (cf-PWV) is the gold standard for measuring aortic stiffness. Finger-toe PWV (ft-PWV) is a simpler noninvasive method for measuring arterial stiffness. Although the validity of the method has been previously assessed, its accuracy can be improved. ft-PWV is determined on the basis of a patented height chart for the distance and the pulse transit time (PTT) between the finger and the toe pulpar arteries signals (ft-PTT).Method:The objective of the first study, performed in 66 patients, was to compare different algorithms (intersecting tangents, maximum of the second derivative, 10% threshold and cross-correlation) for determining the foot of the arterial pulse wave, thus the ft-PTT. The objective of the second study, performed in 101 patients, was to investigate different signal processing chains to improve the concordance of ft-PWV with the gold-standard cf-PWV. Finger-toe PWV (ft-PWV) was calculated using the four algorithms.Results:The best correlations relating ft-PWV and cf-PWV, and relating ft-PTT and carotid-femoral PTT were obtained with the maximum of the second derivative algorithm [PWV: r(2)=0.56, P<0.0001, root mean square error (RMSE)=0.9m/s; PTT: r(2)=0.61, P<0.001, RMSE=12ms]. The three other algorithms showed lower correlations. The correlation between ft-PTT and carotid-femoral PTT further improved (r(2)=0.81, P<0.0001, RMSE=5.4ms) when the maximum of the second derivative algorithm was combined with an optimized signal processing chain.Conclusion:Selecting the maximum of the second derivative algorithm for detecting the foot of the pressure waveform, and combining it with an optimized signal processing chain, improved the accuracy of ft-PWV measurement in the current population sample. Thus, it makes ft-PWV very promising for the simple noninvasive determination of aortic stiffness in clinical practice.