Customer Reviews

Nonlinear Dynamic Modeling of Physiological Systems (IEEE Press Series on Biomedical Engineering)

5 star:		(2)
4 star:		(0)
3 star:		(1)
2 star:		(0)
1 star:		(0)

 

 

The book gives an introduction, as well as a more in-depth look on a number of parametric and nonparametric modeling approaches, as well as their applications to modern physiological system modeling, focusing on the nonlinearies and nonstationarities that need to be addressed during this approach. Although one must be equipped with sufficient understanding of signal processing and a strong mathematic background to access the field of nonlinear modeling, this book is as accessible as it can get, providing constantly examples and applications for each concept it introduces.

It is a great book for those interested in effective methodological tools for mathematical modeling of nonlinear physiological systems. The book begins by presenting an overview of the physiological system modeling problem with several examples (e.g. vertebrate retina, invertebrate photoreceptor, glucose-insulin minimal model). The author was personally involved in the investigation of many of these models, which makes the presentation and interpretation of the obtained results particularly interesting. This is followed by presentation of the concept of a functional as a mathematical mapping of the past and present values of the input signal x(t) onto the present value of the output signal y(t). This functional representation using input-output physiological data is used for the nonparametric dynamic modeling of nonlinear physiological systems, based on the Volterra functional expansion and its variants (e.g. orthogonal Wiener series). The advantages and limitations of the different models are discussed. The Laguerre expansion technique for Volterra kernel estimation is then presented as a means of significantly reducing the total number of parameters to be estimated, which has the advantage of not only reducing the computational burden of the identification process, but also allowing the use of shorter data sets in the identification process. This is especially interesting when dealing with human physiological data, as is my case, since it allows the possibility of reducing the duration of a particular study and still obtaining meaningful results.

The book focuses on the interpretation of the estimated kernels obtained throughout the book, in particular the first- and second-order kernels, as a means of obtaining dynamic information of the system being studied (e.g. the first order kernels in the time domain are related to the dynamic relationships between past values of the input(s) and the linear component of the output, while the second order kernels investigate the dynamic inter-relationships among pairs of input variables to the quadratic component of the system output). While this first part of the book is my personal main interest at this time (my current work involves the identification of physiologically meaningful parameters in a minimal cardiorespiratory model using the Laguerre expansion technique and its variants), the book also discusses parametric (including the NARMAX model), modular (including the principal dynamic modes of a system, equivalent to the discrete Volterra model), and connectionist models, as well as the modeling of multiinput/multioutput systems and closed-loop systems.

I would recommend this book to anyone interested in a thorough understanding of the identification process of nonlinear physiological models.

Really. Highly recommended. For getting started in this field.

But some of the hyperbole is rather off-putting.