The interest of this paper is in the tracking of rhythmic biomedical signals from electronic sensor systems. The known and noisy biomedical signals and the underlying unknown clinical features like the frequency and phase can be modelled using afirst order hidden Markov state space model. It is known the Bayesian filtering is the most widely used solution for track such models. The Kalman filter is known to be a powerful Bayesian estimator. However, it is limited to linear and Gaussian systems.The particle filter, on the other hand, can be applied to a general class of nonlinear non - Gaussian systems. However, the filter involves high computational complexity due to resampling and the need to guide particles into regions that are important to the posterior probability density. This paper proposes a Monte Carlo based Kalman filter, which is a mixed implementation of the Kalman filter and the particle filter, to effectively track biomedical signals. The simulation results show that the proposed methodis superior both in terms of speed and tracking accuracy than the conventional methods.