A first approach on fault detection and isolation for cardiovascular anomalies detection

In this paper, we use an extended version of the cardiovascular system's state space model presented by [1] and propose a fault detection and isolation methodology to study the problem of detecting cardiovascular anomalies that can originate from variations in physiological parameters and deviations in the performance of the heart's mitral and aortic valves. An observer-based approach is discussed as the basis of the method. The approach contemplates a bank of Extended Kalman Filters to achieve joint estimation of the model's states and parameters and to detect malfunctions in the valves' performance. © 2015 American Automatic Control Council.