In this paper, we present a computationally efficient method for adaptive tracking of physiological parameters such as heart rate and respiratory rate from the arterial blood pressure (ABP) measurement using particle filters. A previously reported estimation and tracking method was based on approximating the nonlinear models to linear ones based on the extended Kalman filters. However, the dynamic state-space model of the time-varying parameters and the ABP measurement is highly nonlinear in nature. In addition, the periodic nature of many of the time-varying parameters tend to make the estimation and tracking problem ill posed. In this light, the Rao-Blackwellized particle filtering method is proposed to adaptively estimate and track those parameters. The Rao-Blackwellized particle filter is capable of estimating the time-varying parameters of a nonlinear state-space model without performing any linear approximations while being computationally efficient. We demonstrate the performance improvements of our proposed method through computer simulations.
2011 IEEE International Symposium on Medical Measurements and Applications, MeMeA 2011
Department of Systems and Computer Engineering

Balasingam, B. (B.), Forouzanfar, M. (M.), Bolic, M. (M.), Dajani, H. (H.), Groza, V. (V.), & Rajan, S. (2011). Arterial blood pressure parameter estimation and tracking using particle filters. In MeMeA 2011 - 2011 IEEE International Symposium on Medical Measurements and Applications, Proceedings. doi:10.1109/MeMeA.2011.5966739