Blind Identification of 2-Channel IIR Wave Propagation Systems for Central Cardiovascular Monitoring

This paper presents a new approach to blind identification of a class of 2-channel infinite impulse response (IIR) systems describing the wave propagation dynamics. For these systems, this paper derives a blind identifiability condition and develops a blind identification algorithm, which is capable of uniquely determining both the numerator and denominator polynomials of the channel dynamics. The efficacy of the method is illustrated by a 2-sensor central cardiovascular monitoring application as an example, where the cardiovascular blood pressure wave propagation dynamics is identified and the aortic signals are reconstructed from blood pressure measurements at two distinct extremity locations. Experimental results using a swine subject illustrate how the new blind identification approach effectively identifies cardiovascular dynamics and reconstructs the aortic blood pressure and flow signals very accurately from two distinct peripheral blood pressure measurements under diverse physiologic conditions.