Comparison of various methods for the estimation for aortic characteristic impedance in time domain from velocity-encoded magnetic resonance and applanation tonometry data

Abstract

Objectives. Tocombinecardiovascularmagneticresonance(CMR),whichprovidesaccurateandnon-invasiveevaluationofflow,withapplanationtonometry toassessaortic characteristic impedance (Zc) in time domain while comparing six different methods. Methods. We studied 72 healthy volunteers (31women, age: 42.1 15.3years) who underwent aortic velocity-encoded CMR and carotid applanation tonometry. We developed a method for the superimposition of flow and pressure waveforms as well as semi-automated estimation of temporal Zc based either on magnitude, derivative peaks and early systolic up-slopes ratios or pressure-flow loop. All Zc indices were used to calculate reflection magnitude (RM) as the ratio between backward and forward pressures. Results. For all methods, Zc was in good agreement with the reference Zc provided in frequency domain and with the theoretical water-hammer Zc, which combines aortic pulse wave velocity and area, and with carotid pulse pressure (table) with a slight superiority for methods based on derivatives peaks and early systolic up-slopes. In addition, only these latter two methods were significantly related to arterial stiffness indices such as tonometric carotid-femoral pulse wave velocity (rZ 0.27 and rZ 0.25; p < 0.03) and CMR ascending aorta distensibility (r Z 0.30 and r Z 0.25; p < 0.03). Again these two methods were slightly superior when comparing the derived RM against age (r Z 0.68, r Z 0.65; p < 0.0001). Conclusions. The time derivative and up-slopes straightforward computation methods that can be easily integrated in a clinical workflow provides