Mean Perfusion Pressure Estimation Using Hemodynamics Ohm's Law Towards Unobtrusive Mean Arterial Pressure Estimation

This work proposes a novel approach to estimating mean perfusion pressure, a difference between mean arterial and central venous pressure, using a pulsatile arterial blood pressure and a flow waveform. Using these waveforms, which can be measured, for example, using ultrasound, a vascular resistance downstream at the branch of an arterial tree is estimated. This estimation exploits Ohm's law in hemodynamics during multiple steady states when beat-to-beat variations in arterial pressure occur physiologically. Then, the mean perfusion pressure is estimated by multiplying the vascular resistance with time-averaged flow. A numerical simulation of hemodynamics in a human arterial tree demonstrates the approach. MGHDB records in PhysioNet were used to emulate natural variations of the arterial pressure. The mean and standard deviation of errors in validation are -0.1 and 0.3 mmHg, respectively. This work establishes a theoretical background towards mean arterial pressure estimation and supplements prior works to produce an absolute arterial blood pressure waveform unobtrusively and without calibration.